--- Log opened Tue Apr 10 00:00:30 2018 --- Day changed Tue Apr 10 2018 00:00 < mawk> my failing server is back online, but now I've got 7 days to move the data from the new server they gave me 00:00 < CodeBug> Sitri DM? 00:00 < blackflag_bfp> CodeBug: My goal one day is to LFS (Linux from scratch) 00:00 < Sitri> DM in this context means...? 00:00 < mawk> we say pm CodeBug 00:00 < mawk> private message 00:00 < vlt> tycoon177: That's what I use to keep files on a dev server in sync with local developemnt machine ;-) 00:01 < CodeBug> mawk DM is direct message same as pm 00:01 < CodeBug> seriously? 00:01 < CodeBug> grammar nazi. 00:01 < mawk> lol 00:01 < mawk> see, Sitri didn't understand you 00:01 < mawk> it's not for being picky 00:01 < lukey> CodeBug: ..., Failover with Pacemaker, ... 00:01 < mawk> it's for you to be understood 00:01 < Sitri> You may private message me. 00:01 < TJ-> krzee: try "sudo losetup -fP --show ~k/Desktop/marino.bin" 00:01 < CodeBug> lol I know that 00:02 < Dagmar> If RHSE then you have to be able to do all of that (except setting up an LDAP server and a load-balancer) in about two hours 00:02 < Dagmar> er RHCE 00:02 < Dagmar> ...plus Samba and firewalling 00:02 < mawk> that seems too little time 00:02 < Dagmar> I agree with that assessment 00:02 < Dagmar> There's not really enough time to actually *check* anything 00:02 < mawk> yeah 00:03 < mawk> I'm always slow when I code/do maths/do some sysadmin because I check and recheck everything, test every corner case, etc 00:03 < Dagmar> I think that's a _bad_ situation to be creating where people are basically expected to and encouraged to just throw configurations out there without testing 00:03 < mawk> yeah 00:03 < Dagmar> The RHCSA is a doddle in the time alloted, but the RHCE is kind of questionable 00:04 < krzee> TJ-: i had not. just ran your command and it outputted /dev/loop0 and mounted the partitions that normally mount when i insert the original sd 00:05 < jkemppainen_> I see we're discussing certs ... anyone here familiar with CompTIA's Linux+? 00:06 < Dagmar> I'm kinda leaning towards LFCS/LFCE lately since they can be done remotely 00:07 < Dagmar> I'd argue that if you can't work remotely then you aren't going to get far as a sysadmin 00:07 < pawnshop> jkemppainen: Worth slightly less than the digital paper it is printed on. But you can use the material structure to guide self study 00:08 < Dagmar> Also the idea of driving four hours to Atlanta (the nearest testing location) and back after 6-8 hours of tests is kind of a non-starter 00:08 < pawnshop> But that goes for a good chunk of certs 00:08 < TJ-> krzee: OK, so now if you do "lsblk /dev/loop0" you see all the execpted partitions? 00:08 < xamithan> linux+ is a waste in US, no one recognizes it. Just do redhat 00:08 < Dagmar> RH requires you go to one of the few testing locations 00:08 < Dagmar> That ain't exactly funsies 00:09 < the_document> my console is totally garbled when I swich from Xorg to some tty using ALT+CTRL+3 00:09 < the_document> how to fix? 00:09 < xamithan> They go through pearson, can't be any worse than comptia who goes through the same company 00:09 < mawk> try ctrl-c the_document 00:09 < mawk> are you really required to have this certification thing in the US ? 00:09 < mawk> to get a sysadmin job 00:09 < Dagmar> So long as it doesn't involve trick questions involving useradd/adduser 00:09 < the_document> mawk: tried 00:09 < the_document> I also tried reset still no help 00:09 < Dagmar> mawk: No, but it helps because HR people are stupid 00:09 < twainwek> the_document: what distro? i saw someone complaining about a weird bug like that recently 00:09 < mawk> I always confuse the two Dagmar 00:10 < mawk> I see 00:10 < xamithan> You aren't required to have any certs unless you government, and all they want is comptia 00:10 < Dagmar> mawk: I haven't willingly used either of them in yers 00:10 < the_document> twainwek: gentoo 00:10 < krzee> TJ-: yes 00:10 < lupine> certification is not useful 00:10 < twainwek> ah nevermind, maybe something different 00:10 < lupine> well 00:10 < TJ-> krzee: It might be the target SD MMC card needs to be zero-wiped to allow the controller to do some remapping. You could try "sudo dd if=/dev/zero of=/dev/mmcblk0 bs=24M status=progress; sync" 00:10 < lupine> it's useful for management to select docile hires 00:10 < the_document> its wierd because its random, sometimes its not garbled 00:10 < lupine> but otherwise, no 00:10 < xamithan> I agree, RHCE hasn't done jack for me 00:10 < mawk> try to restart the correct getty the_document 00:10 < Dagmar> mawk: I just bang out the changes to passwd/group/shadow by hand with vim 00:10 * the_document has does not know how to restart getty 00:10 < mawk> lol 00:11 < lupine> just kill it 00:11 < lupine> it'll probably come back 00:11 < the_document> killall -9 getty? 00:11 < lupine> sure 00:11 < Dagmar> Perhaps not killall 00:11 < mawk> I wouldn't do that but you can try 00:11 < mawk> and perhaps not -9 00:11 < Dagmar> init should restart a murdered getty 00:11 < mawk> fine then 00:11 < lupine> obviously, depends on your init system 00:11 < the_document> oww you meant agetty? 00:11 < mawk> yeah 00:12 < lupine> if you're running some junk like sysvinit you might be out of luck 00:12 < TJ-> krzee: I had a 32GB SD-card in a RasPi that 'failed' and PC couldn't read it (I/O errors) via a USB<>SD-card adapter, but could be accessed via a proper MMC host controller, which allowed me to zero it. Did 36 hours of badblocks tests on it with 100% success and it's back in service 00:12 < the_document> need root 00:12 < mawk> of course 00:12 < Dagmar> No, sysv init will restart those just fine 00:12 < the_document> nope 00:12 < the_document> killed with root 00:12 < the_document> nada 00:12 < treaki> hi 00:12 < the_document> still garbled 00:12 < mawk> unload the tty and reload it the_document 00:13 < mawk> you can do C? 00:13 < treaki> i allready asked in #debian on freenode and oftc 00:13 < mawk> I may have a program that does it lying around, let me pack it 00:13 < treaki> but no ideas 00:13 < mawk> python can do it too by the way, it should be easier actually 00:13 < treaki> i need to debootstrap some debian from a crappy live system 00:13 < the_document> mawk: I held down CTRL+C for like 2 seconds didn't help 00:13 < Dagmar> Meh if you're root you can manually restart the agetty instance 00:13 < krzee> i think we're getting somewhere... now its giving me an error writing to the device (i wasnt getting this earlier) dd: error writing '/dev/mmcblk0': No space left on device 00:13 < mawk> that's not unloading the tty the_document 00:13 < treaki> the problem i have is that it just quits without any notice before the system is finished 00:13 < Dagmar> It'd be an ugly tempfix but it would work 00:14 < TJ-> treaki: what's the problem then? 00:14 < treaki> here are some infos: 00:14 < the_document> mawk: how do I reload a tty 00:14 < treaki> http://paste.debian.net/1019471/ 00:14 < treaki> http://paste.debian.net/1019472 00:14 < treaki> http://paste.debian.net/1019473 00:14 < SporkWitch> don't spam 00:14 < treaki> and the access to my test enviorment 00:14 < mawk> but actually you can't unload the tty if getty is on it the_document so nevermind 00:14 < treaki> have a look at here: http://79.223.105.127:6080/vnc.html?host=79.223.105.127&port=6080&password=viewonly 00:14 < nobrain> post it again, i was unable to read 00:14 < mawk> just try the getty thing the proper way 00:14 < mawk> using systemd the_document ? 00:14 < treaki> i am not spamming, that are just 4 links 00:14 < Dagmar> the_document: Why haven't you just typed `stty sane` or `reset` in that window 00:14 < the_document> openrc 00:14 < spammcoin> spam!!!! 00:14 < the_document> I TRIED RESET 00:14 < krzee> TJ-: actually, when i try to write to mmcblk0 i get that error, but when i write to it's p1 it lets me 0 it out 00:14 < spammcoin> i lovvveeee spam 00:15 < mawk> can't help you then 00:15 < the_document> but it has cholesterol 00:15 < mawk> restart the correct getty the proper way 00:15 < treaki> i of cause could create another paste to link the other pastes but somewhere is the edge between usability and minimalism reached... 00:15 < treaki> i think... 00:15 < the_document> spammcoin: is spam spreadable? 00:15 < spammcoin> you have to blend it first 00:15 < krzee> TJ-: this sounds like its what you thought it was at first... my device wont read mmcblk0 so i gotta write each partition individually 00:15 < ghulan> It's a trilinquint. 00:15 < krzee> wont WRITE to* 00:15 < spammcoin> make a spammshake 00:16 < treaki> anyone here who knows debootstrap?? 00:16 < ghulan> There's only 3 while 6 is a crowd. 00:16 < the_document> isn't it processed meats? another words they mix and match random animal meats together 00:16 < lukey> krzee: try wipefs-a on the destination sdcard an then reinsert it 00:16 < lukey> krzee: wipefs -a 00:16 < spammcoin> smeat(tm) 00:17 < treaki> or are you just able to trot out spam topics ;) ; no offense 00:17 < the_document> Dagmar: stty sane and reset dont help 00:17 < spammcoin> deboot is shutdown ? 00:17 < TJ-> krzee: are you sure the write-protect notch hasn't been accidentially triggered on the card? I've seen that do strange things! 00:17 < krzee> lukey: im still writing 0's to it 00:17 < the_document> its not garbled really, its just letters mismached 00:18 < the_document> rearanged 00:18 < lukey> krzee: are you sure p1 does even exist? 00:18 < TJ-> treaki: when you did the debootstrap --foreign that infers the host architecture is different from that target. Is the host running a 64-bit userpsace (amd64) "dpkg --print-architecture" 00:18 < krzee> TJ-: yes i checked the write tab, and when i write the 0's to p1 it works fine 00:18 < krzee> lukey: umm yes lol 00:19 < TJ-> krzee: that seems weird, you should definitely be able to write to the raw device without partitions 00:19 < treaki> TJ-, good idea, but there are both ix86 systems 00:19 < Psi-Jack> CodeBug: You should stop PMing people randomly. Use the channel. 00:20 < treaki> and i tryed already that parameter... 00:20 < treaki> no difference 00:20 < krzee> im 6.5 gigs out of 8 into the /dev/zero write to /dev/mmcbblk0p1 00:20 < the_document> linux isnt safe https://wikileaks.org/ciav7p1/ 00:20 < spammcoin> lol what's this about 00:20 < TJ-> krzee: check the block device config with: "for n in /sys/block/mmcblk0*/*; do [ -f $n ] && echo $n=$(cat $n); done " 00:20 < mawk> the_document: sudo python3 -c 'import os, fcntl, sys; fd = os.open("/dev/tty0", os.O_NOCTTY | os.O_RDWR); tty = int(sys.argv[1]); fcntl.ioctl(fd, 0x5608, tty); fcntl.ioctl(fd, 0x5606, tty)' 3 00:20 < the_document> those organisations recruit the best of the best to make exploits 00:20 < treaki> the thing is its downloading and validating everything and then it quits before the first I: Extracting .... process is done 00:21 < mawk> that resets the tty, but the getty on it needs to be shut down 00:21 < treaki> so it lookes like there is a problem extracting... 00:21 < TJ-> treaki: WHICH ARCHITECTURE is the host!? 00:21 < spammcoin> x86_64 exploits? 00:21 < krzee> smalllaptop vpns # dd if=/dev/zero of=/dev/mmcblk0p1 bs=24M status=progress; sync 00:21 < krzee> 7927234560 bytes (7.9 GB, 7.4 GiB) copied, 417.308 s, 19.0 MB/s 00:21 < krzee> dd: error writing '/dev/mmcblk0p1': No space left on device 00:21 < krzee> i stand corrected 00:21 < Psi-Jack> the_document: Air isn't safe, because https://www.nytimes.com/2018/04/08/world/middleeast/syria-chemical-attack-ghouta.html 00:22 * treaki is now trying --extractor=ar 00:22 < treaki> ... 00:22 < the_document> mawk: nope https://paste.pound-python.org/show/JAhbiCCZwO7kDLfb3Izq/ 00:22 < TJ-> treaki: if you don't answer specific questions you won't get help 00:22 < mawk> what did I just say the_document 00:22 < mawk> "but the getty on it needs to be shut down" 00:22 < the_document> yea I was reading 00:22 < the_document> I jumped the gun 00:22 < the_document> using browser irc client 00:23 < the_document> Psi-Jack: different matter all together 00:23 < the_document> Psi-Jack: go to #air 00:23 < treaki> yey its working now... 00:23 < Psi-Jack> the_document: Stop trolling in ##linux. Simpler. :) 00:23 < the_document> im not trolling 00:24 < krzee> TJ-: output from for loop: https://paste.linux.community/view/b50de610 00:24 < Psi-Jack> the_document: Join a Linux channel, tell everyone in the channel their system is not safe. How is that NOT trolling? :p 00:24 < mawk> by the way the_document you can put what's -c '>here<' into a .py script with #!/usr/bin/env python3 on the first line, then you can call it like ./reset.py 3 00:24 < nolsen> Psi-Jack: Wow, I think I found the problem... 00:24 < treaki> now its installing the base system ;) 00:24 < nolsen> I think it was dust buildup on the contact pins 00:24 < Psi-Jack> nolsen: YOu were using USB 3.0? :P 00:24 < nolsen> Nope, it was dust buildup 00:25 < TJ-> krzee: "/sys/block/mmcblk0/ro=1" ?? RO == READ ONLY 00:25 < nolsen> and USB 3.0 isn't horrible. 00:25 < Psi-Jack> I highly doubt it. :p 00:25 < treaki> thx a lot 00:25 < nolsen> I unplugged it, blow out the dust on both connectors, then plug it back in 00:25 < Psi-Jack> treaki: "thanks" not "thx" for future self corrections. 00:25 < nolsen> then boom, 70MB/s 00:25 < the_document> uhh I thought ingnore blocks a users chat 00:25 < dviola> I've fixed my mech keyboard, I forgot how good it was to type on one 00:26 < CodeBug> Psi-Jack, are you a grammar nazi? 00:26 < the_document> dviola: you actually like the clinking sound? 00:26 < Psi-Jack> CodeBug: You seem to not understand what grammar actually is. 00:26 < nolsen> Psi-Jack: You don't need a hardware card for raid. 00:26 < nolsen> USB 3.0 is great. 00:26 < krzee> TJ-: palm ---> face (flipped tab and trying again, if that doesnt work i will grab another mini to full size sd adapter) 00:26 < ayecee> nothing so severe. more like a grammar aficiendo. 00:26 < dviola> the_document: it doesn't bother me, but I think it can be annoyign to others 00:26 < dviola> annoying* 00:27 < krzee> currently running the dd if=/dev/zero to see if i can do that without issue 00:27 < the_document> if only they made mech boards which DONT have that noise... 00:27 < dviola> the_document: why don't you like it? 00:27 < TJ-> krzee: once the card is in, rerun the 'for' look and check the reported value. Some host controllers don't detect it properly so it might have been a 'fake' value 00:27 < krzee> TJ-: it indeed changed to 0 00:28 < the_document> dviola: its like positive feedback loop which eventually causes me to search and buy a different keyboard 00:28 < TJ-> krzee: so, sanity is restored! 00:28 < blackflag_bfp> n00b time is there a keystroke for jumoing to specific buffer? Not finding it in the man 00:28 < TJ-> krzee: the card's internal MMC controller was accepting the data but discarding it 00:28 < blackflag_bfp> oops found it 00:28 < treaki> blackflag_bfp, on which viewer/editor? 00:28 < blackflag_bfp> Shift + # 00:28 < blackflag_bfp> weechat 00:28 < treaki> ah 00:29 < the_document> mawk: this any good? https://paste.pound-python.org/show/tNfW5YDwzB5YG27brCkj/ 00:29 < TJ-> blackflag_bfp: or Alt+J 00:29 < blackflag_bfp> treaki: thanks 00:29 < nobrain> F5/F6 prev/next 00:29 < TJ-> blackflag_bfp: or for last active Alt+A 00:29 < krzee> TJ-: i swear i had tried the original command with the tab in both positions, so i *hope* that was the problem but i have a feeling i'll run into something more 00:29 < sammm> hey guys, I want to set up an exchange (office 365) mailbox on a box, it will only receive emails with a certain attachement. what's the best method to set this up? 00:29 < the_document> i want to replace tty0 with tty* 00:29 < treaki> np, 00:29 < blackflag_bfp> TJ-: thanks! 00:30 < mawk> no the_document like this https://paste.pound-python.org/show/qPU59YthoX5mQWB3uzPH/ 00:30 < mawk> don't replace tty0 00:30 * the_document notices _systemd_is_evil nick 00:30 < dviola> the_document: I see 00:30 < blackflag_bfp> nobrain: thanks! 00:30 < mawk> you run it like `./reset.py 3' for tty3 for instance 00:31 < TJ-> blackflag_bfp: in the weechat buffer "/key list" to see all mappings 00:31 < Psi-Jack> the_document: Yeah. I'm betting that's a idle-bot, never says anything. 00:31 < the_document> Psi-Jack: parked nickname 00:32 < blackflag_bfp> TJ-: Super Gold Star for you brother! Good Intel 00:32 < the_document> mawk: ehh https://paste.pound-python.org/show/9HmxpT2pBfS6Fq2CdAPN/ I don't use .py extension but who cares anyways 00:32 < mawk> what I said about getty still stands the_document 00:32 < mawk> it won't deallocate the vt until nobody's on it anymore 00:33 < mawk> that's how it works 00:33 < the_document> oww 00:33 * blackflag_bfp is lurking about watching all the sweet information tick by 00:33 < mawk> tell openrc to stop the getty on tty3 00:33 < Alexander-47u> has libreoffice always been this buggy? 00:33 < phogg> Alexander-47u: is it buggy now? 00:33 < Alexander-47u> for me it is 00:33 < the_document> mawk: I thought I had to do it onto a tty which wasen't logged into 00:33 < MrElendig> Alexander-47u: yes 00:33 < krzee> TJ-: now with the for loop i get ro: 0 but i still get the same no space left on device error... im going to try another micro to full sd adapter 00:33 < Alexander-47u> small documents are fine 00:34 < MrElendig> Alexander-47u: fresh is usually better in my experience btw 00:34 < Alexander-47u> but above 100 pages, it just goes apeshit 00:34 < phogg> Alexander-47u: There have always been some compatibility issues, and crashes are not impossible. It used to be worse when half of it was java. 00:34 < mawk> even if you're not logged in, the getty program is still running on it the_document 00:34 < TJ-> krzee: that report is expected! 00:34 * the_document disagrees with Alexander-47u 00:34 < mawk> the getty program and the login program 00:34 < TJ-> krzee: you're running dd with no limit so it keeps going until there's no space left! 00:34 * the_document likes ebooks 00:34 < phogg> Alexander-47u: that does not match my experience. I've opened pretty huge documents without trouble. 00:34 < krzee> ohhhh LOL good then 00:34 < phogg> Alexander-47u: lowriter? 00:34 < the_document> lol 00:35 < the_document> mawk: agetty isn't even in any runlevel 00:35 < the_document> mawk: it wasen't even running 00:35 < mawk> it's running, otherwise you wouldn't have login prompts on the ttys 00:35 < the_document> hmm 00:35 * the_document stops agetty 00:36 < Alexander-47u> phogg, i get troubles with the layout 00:36 < Alexander-47u> for example 00:36 < TJ-> krzee: when dd finishes do "sync" to ensure all data has been flushed from kernel cache into the card before you remove it 00:36 < Alexander-47u> if I scroll through too fast, images just move by itself to the next page, messing up the layout 00:36 < the_document> mawk: your script is buggy 00:37 < phogg> Alexander-47u: is there a publicly accessible document that has this problem for you so I can try and see what you're seeing? 00:37 < the_document> tried running it with and without agetty service active 00:37 < the_document> but whatever 00:37 < Alexander-47u> phogg, not that i know of, and this is for my course so I cant share 00:37 < the_document> it only happens if I uses Xorg 00:38 < Alexander-47u> no worries, just thought I'd express my minor frustration with it 00:38 < Alexander-47u> it still works, just have to check if the layout has not messed up by itself quite frequently 00:38 < the_document> is it true systemd resolved a lot of init script headaches at the cost of complexity and implimentating a windows style app into linux? 00:39 < Alexander-47u> maybe its because of many screenshots includs 00:39 < the_document> thats what I understand it did 00:39 < Alexander-47u> included 00:39 < spammcoin> the_document: idk, never used it 00:39 < the_document> spammcoin: you probably did, if you used some popular distros livecd 00:40 < spammcoin> the_document: i never used it. 00:40 < the_document> what distro u running? 00:40 < mawk> it's not buggy the_document 00:40 < spammcoin> not your grandmothers linux (patent pending) 00:40 < mawk> the script is doing only one thing, calling the ioctl 00:40 < mawk> you're seeing the raw ioctl response 00:41 < mawk> if it doesn't want to reload the tty, either the tty is already unloaded either there is still a process on it 00:42 < slidinghorn> I don't know if I should feel dumb, or if this is just legit crap documentation.. (probably the former) - looking at irssi's page trying to figure out what their "tip/trick" for right-aligned nicks *actually means* 00:42 < the_document> mawk: ps shows login - bash and ps ... 00:43 < krzee> TJ-: after doing the dd to thee sd card i still only have 1 partition on it 00:43 < MrElendig> slidinghorn: just be glad they have any documentation at all now, until recently it was all "to be written" 00:43 < MrElendig> :p 00:43 < the_document> krzee: you sure your image was a multi partition image? 00:43 < RayTracer> slidinghorn: maybe its about right-to-left languages? 00:43 < mawk> the_document: sudo ps -f -t tty3 00:44 < slidinghorn> RayTracer: I know what it's supposed to do, I just want to adjust its behavior, but I can't figure out what the hell any of what it says means... 00:44 < krzee> the_document: yep, as seen when i run the losetup command that TJ-gave me to run above 00:45 < TJ-> krzee: have you rescanned the device? 00:45 < TJ-> krzee: "sudo partprobe /dev/mmcblk0" 00:45 < RayTracer> slidinghorn: ok maybe I'm not the best guide here, as I don't use irssi (and in consequence don't have its man page) ;) 00:45 < krzee> TJ-: Error: Can't have a partition outside the disk! 00:46 < the_document> krzee: never done dd imaging with multi partitions, I should play around with the concept 00:46 < krzee> i cant believe how much issue im having with such a simple process lol 00:46 < TJ-> krzee: can you show me "sudo hexdump -n 16384 /dev/mmcblk0 | nc termbin.com 9999" 00:47 < neachdainn> What could cause a TCP connection to timeout if both ends are local? 00:47 < the_document> mawk: yea it says bash and login 00:47 < mawk> then something is on the tty the_document ... 00:47 < krzee> TJ-: http://termbin.com/x2nm 00:47 < mawk> kill everything in it, stay out of it, start my script 00:47 < mawk> and there you go 00:47 < mawk> reloaded tty 00:48 < mawk> if it's still garbled I can't do much more 00:48 < krzee> neachdainn: firewall, bad cable, bad route, bad switch port... 00:48 < the_document> i bet I should start kde with & suffix 00:48 < the_document> so it wont write to term 00:48 < mawk> what ? 00:48 < mawk> it will still write to term by the way 00:48 < TJ-> krzee: well there is a partition table in there... is this SD-card smaller than the original SD-card ? 00:48 < the_document> mawk: the xinitrc 00:48 < mawk> if you want to start it as a daemon you can do: yourcommand >/dev/null 2>& & disown 00:48 < krzee> TJ-: same size from the same shipment, same brand and everything even 00:49 < mawk> sorry: yourcommand >/dev/null 2>&1 & disown 00:49 < neachdainn> krzee: Local, as in on the same machine. It should never go through a switch, cable, or firewall. 00:49 < the_document> mawk: yea thats optimal, will modify my xinitrc 00:49 < TJ-> krzee: in the earlier 'for' loop output what was /sys/block/mmcblk0/size= ? 00:49 < the_document> mawk: thanks 00:49 < mawk> wait the_document , how do you start the xinitrc ? 00:49 < TJ-> krzee: compare that with what "sudo fdisk -l /dev/mmcblk0" reports 00:49 < krzee> neachdainn: oh i see. could still be firewall though 00:49 < mawk> I don't think you're supposed to do that inside the xinitrc itself 00:50 < mawk> but I've never written one 00:50 < the_document> mawk: startx 00:50 < the_document> mawk: perhaps I need startx >/dev/null 2>&1 & disown 00:50 < mawk> yes 00:50 < krzee> o.O this is interesting 00:50 < neachdainn> krzee: I guess maybe. What's weird really weird about this though is that the connection works for a while and then stops. 00:50 < neachdainn> It used to work for months, then weeks, then days, now I'm lucky to get a few hours 00:51 < the_document> mawk: cant i make a script with startx >/dev/null 2>&1 & disown ?? 00:51 < krzee> TJ-: check this out http://termbin.com/v00w 00:51 < mawk> you can the_document 00:51 < krzee> thats an 8gb chip TJ- 00:51 < mawk> but every sane distro already provides one normally 00:51 < the_document> mawk: I guess disowning the name of the script is easier than writing out >/dev/null 2>&1 & disown 00:51 < TJ-> neachdainn: what processes are using this TCP connection? 00:52 < mawk> disown means don't kill the last &'d process when we exit 00:52 < mawk> you don't disown a command 00:52 < TJ-> krzee: does the sector count match what /sys/block/... size= reports/ 00:52 < the_document> mawk: I thought the script will still linger around 00:52 < the_document> as a process 00:53 < mawk> why ? 00:53 < neachdainn> TJ-: ROS and ROS related processes. 00:53 < treaki> hi there 00:53 < mawk> using the & disown trick the backgrounded process will fly with its own wings 00:53 < the_document> meh but I think problem solved 00:53 < mawk> and the current script will terminate normally 00:53 < treaki> ive got some old binary that is only single threaded 00:53 < treaki> on my ryzon it would be grate to use all avaivalbe cores because it dose havy duty 00:54 < aaro> the_document, in xinitrc you start your de/wm with exec, not & 00:54 < mawk> then modify the code treaki 00:54 < RayTracer> treaki: maybe you can run that binary multiple times 00:54 < TJ-> neachdainn: possibly the kernel socket timeout is being hit, is there a keepalive set? 00:54 < treaki> is there some way to wrap/emulate/whatever do to distribute all that work to all that cores 00:54 < TJ-> treaki: no 00:54 < treaki> no, i cant run it multiple times because it is a linear process 00:54 < mawk> no 00:54 < MrElendig> treaki: only if the data can be processes in independent chunks 00:54 < treaki> its an old game engine ... 00:54 < neachdainn> TJ-: I have no idea. I'll check 00:55 < krzee> TJ-: https://paste.linux.community/view/609d4616 00:55 < treaki> i mean isnt there something like dynamic recompiling or so 00:55 < MrElendig> with game engines you are always stuck with a master thread which will limit the performance 00:55 < treaki> that just checks the asembler and takes all that can be paralized save (independent steps) and spreads them to multiple threads 00:56 < treaki> i mean something like qemu 00:56 < treaki> or so 00:56 < MrElendig> how much can be paralleled varies vastly depending on the game and architecture 00:56 < RayTracer> treaki: if it's an old game, it shouldn't hit any limit on current cpu anyways 00:56 < chindy> I used lxappearance to change the themes of my windows. Now I have a bright theme on all my applications except all KDE based applications, these still have some dark theme. How can I change this ? 00:56 < MrElendig> RayTracer: tell that to supcom 00:56 < MrElendig> or dwarf fortress 00:56 < treaki> a full hardware emulation that dose virtual singel core processor hardware that runs on all cores... 00:56 < MrElendig> or crysis 00:56 < MrElendig> :p 00:56 < AndroidKitKat> is there anyway for a hardware event like closing my laptop lid trigger a program from the command line 00:57 < MrElendig> treaki: that would most likely reduce performance 00:57 < treaki> i mean there are things like ps3/4 nintendo switch and so on emulators 00:57 < treaki> which do things like that 00:57 < MrElendig> AndroidKitKat: multiple ways to do that 00:57 < AndroidKitKat> whats the easiest 00:57 < MrElendig> a bit of a xyproblem though 00:57 < treaki> isnt there some that dose x68 on x68 with dynamic recompiling and optimization? 00:57 < AndroidKitKat> people have told me to try a daemon, but i have no idea what to do then 00:57 < MrElendig> *what* do you want to happen? 00:58 < AndroidKitKat> I want it to trigger `gnome-screensaver-command -l` 00:58 < AndroidKitKat> i3lock is bad imo 00:58 < anklebreaker> what 00:58 < geheimnisse> androidkitkat: acpi events 00:58 < treaki> MrElendig, if i can use quemu and do some software emulation that decreases the performens about 20% (without using kvm) 00:58 < MrElendig> AndroidKitKat: so a total xyproblem 00:59 < treaki> and i put that up to all 12 cores 00:59 < AndroidKitKat> a what? 00:59 < MrElendig> AndroidKitKat: gnome itself can lock on lid close 00:59 < AndroidKitKat> its not gnome 00:59 < AndroidKitKat> its i3 00:59 < treaki> then i have an performance gain 00:59 < geheimnisse> androidkitkat: https://wiki.archlinux.org/index.php/Acpid 00:59 < AndroidKitKat> also not on arch btw 00:59 < treaki> even if i loose some before 00:59 < MrElendig> last I checked gnome-screensaver required gnome-session 00:59 < geheimnisse> archwiki is still informative 00:59 < AndroidKitKat> I'm able to trigger it 00:59 < krzee> TJ-: i would answer directly but im not sure and dont want to answer wrong 00:59 < treaki> like now it job x dose take 42 hours native (one thread only) 01:00 < treaki> then i run it in quemu and it dose 42*1.24 hours 01:00 < neachdainn> TJ-: I think we're good on keepalive. It looks like the real problem might be the TCP socket server going down for some reason. 01:00 < treaki> or so 01:00 < phogg> AndroidKitKat: because arch doesn't auto-configure things it documents how you can do it manually instead, so it's one of those places that has a lot of examples and overviews of what kinds of things you can do. Gentoo's wiki is also good for that. Most of it isn't specific to arch. 01:00 < treaki> and then i run it on all 12 cors (physical 8 so lets add faktor 0.8 to it) 01:00 < MrElendig> mostly due to arch not messing with upstream generally 01:00 < treaki> (42*1.24/12)*0.8 01:01 < treaki> and then the work is done in 3.2 01:01 < AndroidKitKat> yeah phogg just trying to figure it out for my distro is the tricky part 01:01 < TJ-> krzee: that looks weird! the kernel believes the device is 8GB. I tested a similar situation here with a 32GB card and fdisk reports the same as the kernel 01:01 < treaki> even if emulation looses a lot of performance its faster after all MrElendig 01:01 < treaki> do you know such a pice of software? 01:01 < krzee> TJ-: the device *is* 8gb 01:02 < MrElendig> treaki: not going to work that way for a sequential workload 01:02 < RayTracer> krzee: fwiw, you probably can replace that for loop with "grep . /sys/block/mmcblk0*/*" 01:02 < MrElendig> and you are ignoring the cost of context switching etc 01:02 < treaki> MrElendig, not if there are dependencies 01:03 < treaki> but the thing is that the application i have running itself has a lot of multiple threads inside that single thread 01:03 < krzee> RayTracer: ya i noticed it could be changed too, but i ran the command i was told to run as told to run it, not trying to play with the commands im given by those helping me =] 01:03 < MrElendig> that makes no sense at all 01:03 < krzee> dont wanna annoy those nice enough to help me :D 01:03 < treaki> so i guess for a dynamic recompiler that would figure out the dependencies that have to be done linear it could do so 01:04 < treaki> like live migration of a vm 01:04 < treaki> or hot standby failover 01:04 < treaki> techics where on one thread it checks if everything is all right 01:04 < treaki> and if so it takes the faster work 01:04 < treaki> and if not it takes the slower 01:04 < wadadli> just great my keepassx database refuses to open, with all of my credentials locked away in it. 01:04 < treaki> until its save and then it datches the single thread 01:05 < treaki> and distributes all the work to multiple threads 01:05 < blackflag_bfp> time to groll steaks whos coming over? :) 01:06 < TJ-> krzee: right, and fdisk should report that, not the file's image size. 01:06 < blackflag_bfp> and when im done grolling I'll grill them :P 01:07 < TJ-> krzee: RayTracer using grep won't ignore directories and it generates multiline output for some nodes, such as uevent 01:07 < treaki> "Dynamic binary translation" 01:07 < treaki> single to multi thread 01:07 < treaki> such a thing 01:07 < spammcoin> spam 01:08 < RayTracer> TJ-: I see 01:08 < krzee> TJ-: see, thats why i will NOT change the commands you give me ;] 01:08 < treaki> a found something: https://www.researchgate.net/publication/254005131_HQEMU_A_multi-threaded_and_retargetable_dynamic_binary_translator_on_multicores 01:08 < treaki> thanks for your help 01:08 < krzee> hahah 01:08 < TJ-> krzee: the size issue is weird... can the system actually fsck the last partition? "sudo fsck /dev/mmcblk0p7" 01:09 < krzee> fsck.ext2: No such file or directory while trying to open /dev/mmcblk0p7 01:09 < krzee> Possibly non-existent device? 01:09 < krzee> TJ-: no 01:10 < T-Rog> What do I do with an any.pkg file 01:10 < krzee> TJ-: shall i try that with the chip that works?? 01:10 < TJ-> krzee: what brand are these cards/ 01:10 < spammcoin> T-Rog: what kind of file is it, what does `file` report 01:10 < TJ-> krzee: yes, do 01:11 < MrElendig> sidenote: don't use ext2 01:11 < krzee> TJ-: sandisk ultra microsdhc uhs-I 01:11 < spammcoin> why not? 01:11 < TJ-> krzee: I'd have to start suspecting you've got a fake card, or a broken one 01:11 < MrElendig> sure you have 7 partitions on the card? 01:12 < TJ-> MrElendig: Yes 01:12 < MrElendig> what does lsblk / fdisk say about it? 01:13 < krzee> TJ-: ok so first ill toss in the chip that works, ill get fdisk -l and ill try to fsck p7... then i'll give it a shot writing to another card (i have like 10) 01:13 < T-Rog> spammcoin: it's actually a directory. It's a package I downloaded from the AUR 01:13 < T-Rog> I don't have an Ethernet cord and my WiFi is not working 01:13 < T-Rog> I'm trying to install a package offline 01:13 < MrElendig> T-Rog: man pacman 01:14 < TJ-> krzee: That's a good plan. 01:14 < MrElendig> T-Rog: the aur page on the wiki also tells you how to build and install packages 01:14 < MrElendig> T-Rog: what is this package and why don't you have network though? 01:14 < T-Rog> The drivers for my wifi dongle are not installed 01:15 < MrElendig> which dongle 01:15 < MrElendig> (and driver) 01:15 < MrElendig> and are you on the install image or on a installed system 01:15 < T-Rog> I need to install rtl8812au_8821au-linux-dkms I think 01:15 < T-Rog> And it depends on dkms which I don't have installed 01:16 < MrElendig> I suggest trying the rtl8xxxu driver first 01:16 < MrElendig> and please answer that last question 01:16 < T-Rog> It's a Netgear a6100 01:16 < T-Rog> I'm on an installed system 01:16 < krzee> TJ-: fsck works on orig card, fdisk looks same, opening another card now 01:18 < mawk`> I'm inside Haiku's IRC client 01:18 < AndroidKitKat> anyone know how to setup systemd hardware events? ive been reading the wiki and docs but im not quite sure 01:19 < AndroidKitKat> im not quite sure if my system suspends when the lid is closed with i3, although I could be wrong 01:19 < mawk`> don't you have a blinking led or something ? 01:19 < MrElendig> systemd will by default suspend on lid close, unless there is some inhibitor 01:19 < AndroidKitKat> i dont have any leds 01:19 < MrElendig> see systemd-inhibit 01:20 < nai> AndroidKitKat: look into /etc/systemd/logind.conf 01:21 < krzee> TJ-: same problem on new chip, trying with different adapter (that came with new chip) 01:21 < AndroidKitKat> so theyre all commented out, and I guess I just have to figure out on suspend to issue a command 01:21 < nai> AndroidKitKat: what do you want to do exactly? 01:21 < T-Rog> Am I going about this the wrong way? 01:22 < AndroidKitKat> setup my lockscreen with i3 01:22 < AndroidKitKat> one doesnt work ootb 01:22 < krzee> wait wait wait... this is way too fast 01:22 < nai> oh 01:22 < krzee> smalllaptop vpns # dd if=~k/Desktop/marino.bin of=/dev/mmcblk0 bs=4M status=progress; sync 01:22 < krzee> 3825205248 bytes (3.8 GB, 3.6 GiB) copied, 10.007 s, 382 MB/s 01:22 < nai> i don't know then 01:22 < MrElendig> T-Rog: connect a cable or download all the deps onto a usb stick and transfer them 01:22 < mawk`> if it's mostly sparse it could be realistic krzee 01:22 < AndroidKitKat> hmmm 01:22 < MrElendig> T-Rog: but as I said, test the rtl8xxxu driver first 01:22 < mawk`> but I don't know if dd uses sparse by default 01:22 < krzee> oh ok ya i dunno 01:22 < AndroidKitKat> can you see when the last suspend was? 01:23 < T-Rog> I thought I did download the deps but I apparently don't know how to install them 01:23 < RayTracer> krzee: it probably waits on the sync, while the timings are from before sync; you could try oflag=direct 01:23 < nai> AndroidKitKat: i guess what you want is 1/ suspend your computer on lid switch, 2/ show lock screen when going out of suspended mode 01:23 < T-Rog> What I downloaded is an any.pkg.tar.xz file 01:23 < mawk> https://pix.watch/gsBm0w/XSRCqL.png 01:23 < nai> those are 2 separate issues 01:24 < AndroidKitKat> they're different issues D: 01:24 < nai> the first one takes place in logind.conf, you can just put HandleLidSwitch=suspend or something like that 01:24 < krzee> TJ-: on the new chips i cant fsck any partition except for p1, on the original i can 01:24 < nai> as for the lock screen, i don't know, look into i3's documentation 01:24 < AndroidKitKat> thanks nai 01:24 < krzee> TJ-: i can open another chip if you think it may help... 01:25 < TJ-> krzee: you're using some kind of external USB-connected adapter? 01:25 < krzee> no, built in on my dell latitude e7240 01:25 < nai> AndroidKitKat: https://faq.i3wm.org/question/6665/lockscreen-after-closing-my-laptop.1.html 01:25 < AndroidKitKat> gotta find a way around xautolock 01:26 < TJ-> krzee: I thought so; can you "fsck /dev/loop0p7" (the loopdev of the file) ? 01:27 < TJ-> krzee: it's possible the file itself is broken, it'd make more sense 01:27 < RayTracer> krzee: did you compare the original chip with the image on disk? I'm not really deep into your issue, but I suspect the 8015200 sectors total vs. partition to 15523839 is it.. and this was also on the image file 01:27 < TJ-> krzee: also what size is the file itself? if it is an image of an 8GB card it should be 8GB 01:29 < krzee> cannot fdisk the loop device, good call man! 01:29 < TJ-> krzee: the file could be less if it's been written as sparse (that means sectors that were all zero aren't copied, they're marked as empty, to save space) 01:29 < TJ-> krzee: Aha! phew, I was beginning to think you had magic gremlins! 01:29 < krzee> ya its like 4GB 01:29 < TJ-> krzee: "sudo losetup -d /dev/loop0" to drop it, then you can delete the file 01:30 < krzee> deleted 01:31 < TJ-> krzee: then re-copy from the good card. "sudo dd if=/dev/mmcblk0 of=~k/Desktop/marino.bin bs=24M status=progress; sync" 01:31 < TJ-> krzee: once it's done check the file size, and fdisk -l it too 01:32 < krzee> and mount it and fsck that partition! 01:32 < TJ-> krzee: loopdev it yes 01:34 < krzee> smalllaptop vpns # sudo dd if=/dev/mmcblk0 of=~k/Desktop/marino.bin bs=24M status=progress; sync 01:34 < krzee> 4102029312 bytes (4.1 GB, 3.8 GiB) copied, 15.0647 s, 272 MB/s 01:35 < krzee> still 4GB, and still cant fsck those partititons 01:36 < TJ-> krzee: any I/O errors reported via "dmesg" ? 01:36 < krzee> negative ghost rider 01:37 < TJ-> krzee: I'm pretty sure dd won't do a sparse write unless you give it "conv=sparse" option, but retry the dd with that and see if anything differs 01:37 < RayTracer> this partitioning looks like the original chip should be 30gb 01:38 < RayTracer> or maybe I'm too tired to do math.. leaving for bed, good luck to all 01:39 < TJ-> RayTracer: 30GB? it looks like 8GB to me: /sys/block/mmcblk0/size=15523840 01:39 < TJ-> 32GB is 62521344 sectors 01:39 < RayTracer> this strengthens my second theory ;) n8 01:40 < TJ-> LoL 01:40 < krzee> hahah 01:40 < TJ-> g'night 01:40 < krzee> gnite RayTracerthanks man =] 01:40 < krzee> hmm, still 4GB, and it takes 14sec to copy it... seems awfully fast 01:41 < TJ-> krzee: once it's clear, try removing the SD-card then re-running the 'for' loop to see if mmcblk0 is still reported, and if so, what it's size= is 01:41 < TJ-> krzee: kernel caching- maybe we should try with forcing direct access 01:43 < krzee> TJ-: no output from the for loop 01:43 < TJ-> krzee: good, so we're not suffering a static/ghost sysfs entry 01:44 < TJ-> krzee: OK, put the good card back in and let's try another dd operation, command to follow... 01:44 < TJ-> krzee: ... "sudo dd if=/dev/mmcblk0 of=~k/Desktop/marino.bin bs=24M status=progress conv=sparse iflag=direct,dsync,sync" 01:46 < TJ-> krzee: whilst that is running, from another terminal shell, can you do "dmesg | tail -n 200 | nc termbin.com 9999" 01:46 < krzee> smalllaptop vpns # dd if=/dev/mmcblk0 of=~k/Desktop/marino.bin bs=24M status=progress conv=sparse iflag=direct,dsync,sync 01:46 < krzee> dd: failed to open '/dev/mmcblk0': Invalid argument 01:46 < krzee> weird... 01:46 < krzee> that IS the right dev 01:46 < TJ-> krzee: my args are wrong, hang on 01:47 < TJ-> krzee: what does "dd --version" report ? 01:47 < krzee> dd (coreutils) 8.25 01:47 < krzee> Copyright (C) 2016 Free Software Foundation, Inc. 01:47 < krzee> License GPLv3+: GNU GPL version 3 or later . 01:48 < TJ-> krzee: I'm testing with 8.28 but I don't think it's that, I suspect we need to use the oflag not iflag 01:49 < TJ-> krzee: try this "sudo dd if=/dev/mmcblk0 of=~k/Desktop/marino.bin bs=24M status=progress conv=sparse oflag=direct,dsync,sync" 01:49 < krzee> same error 01:49 < wadadli> someone suggest a good pw manager? 01:50 < TJ-> krzee: try dropping the final ",sync" 01:51 < krzee> same error with iflag and oflag 01:51 < TJ-> krzee: ok, let's try something entirely different! 01:52 < krzee> oh but i was slightly wrong too, the filename in the error changes with iflag and oflag 01:52 < krzee> but ya 01:52 < TJ-> krzee: try this "sudo dd if=/dev/mmcblk0 of=~k/Desktop/marino.bin bs=24M status=progress conv=sparse,fdatasync,fsync" 01:53 < AndroidKitKat> I got it working nai 01:53 < nai> nice 01:53 < krzee> TJ-: running, gunna run your dmesg now 01:53 < TJ-> krzee: this command should run much slower - those conv FLAGS make stop it caching in RAM 01:53 < AndroidKitKat> Should I document my solution 01:53 < krzee> damn its over, still FAST 01:53 < krzee> didnt get to tail while running lol 01:54 < TJ-> krzee: that doesn't seem correct, even when writing sparse the read should take some time. Give me the dmesg log anyhow 01:55 < TJ-> krzee: usually, even for class 10 devices, the PC mmc host controllers are quite slow. Maybe you've got a very fast controller though :) 01:55 < TJ-> krzee: do a loopdev fsck on the resulting file too 01:56 < blackflag_bfp> ahh lol it is /away again 01:56 < krzee> TJ-: http://termbin.com/41w5 <-- dmesg|tail output 01:56 * blackflag_bfp is a geezer relearning stuffz 01:59 < TJ-> krzee: Argh! errors. It looks like whenever you're plugging in the SD-card something is auto-mounting the file-systems (possibly the desktop environment via udisks) and then you eject it without unmount/safelty remove, and it is/has caused file-system corruption. Next time you insert it the system then deso fsck's on the partitions 02:01 < krzee> i been ejecting from file manager but there may have been a time i didnt 02:02 < TJ-> krzee: oh yeah, most of the errors are for the loopdev! sorry - tired here 02:03 < krzee> should i go boot it on the pi and let it fix itself if that could be a problem? 02:03 < TJ-> krzee: it won't be the cause of this issue you're seeing, that is just another issue that won't help have valid file-systems if it were interrupted. 02:03 < krzee> TJ-: please dont appologize to me, i really really appreciate all your help / time =] 02:03 < TJ-> krzee: does the fsck of the loopdev still fail? 02:04 < TJ-> krzee: the problem is occuring at the read-from-device stage that's for sure, but when you fsck the device itself it seems fine! 02:06 < TJ-> krzee: for the source card itself can you show me "sudo dumpe2fs -h /dev/mmcblk0p7 | nc termbin.com 9999" 02:08 < krzee> hmm now it doesnt even want to loopdev p7 02:09 < krzee> ill do the dumpe2fs first, then ill make another image the orig way and see if i could loopdev p7 again on that 02:09 < FutureTense> I installed XRDP and when trying to login, I get a blue screen and it just hangs 02:09 < krzee> i get an error when i try to mount the loopdev p7 by clicking on the partition that shows up in finder 02:10 < krzee> http://termbin.com/04ty TJ- 02:11 < krzee> i jjust got let out of work and have to drive home to work more from home, so im sorry but i gotta bbiaf 02:11 < FutureTense> Ah... i just times out and tells me there is a connection problem, giving up 02:11 < krzee> gotta detach from my screen, sorry to bounce in the middle of help TJ- but ill brb in about 20 if thats not too long 02:14 < mawk> what's the best way to read() from a file that has no predefined length in C++ ? 02:14 < TJ-> krzee: OK... it's gone 0100 here so I need to get to bed. dumpe2fs looks fine. Not sure what else to suggest. 02:14 < ayecee> a piece at a time 02:15 < Sitri> In 1kB chunks 02:15 < ayecee> voting 4kb chunks here 02:15 < ayecee> or multiple of that 02:16 < Sitri> Or memmap it then malloc then memcpy. Bam. Entire file in RAM. 02:16 < mawk> I made a std::string and use it as a dynamic container as my buffer for read() 02:16 < mawk> resizing the string by doubling its length each time it's full 02:17 < mawk> but then I'm overcoming the dynamic resing algorithm, I know it's exponential so I 02:17 < phinxy> Could tar -cf change permissions of files if it were a root filesystem? 02:17 < mawk> so I musn't be too far with my doubling every time* 02:17 < mawk> yeah I thought of that as well ayecee 02:18 < lektronx> what is the regex to sort filenames with a numerical suffix? eg access.log.[1-20]? 02:18 < mawk> like mmap'ing it like Sitri said 02:18 < kurahaupo_> phinxy: tar sets permissions on extraction 02:18 < phinxy> doh 02:18 < mawk> but I don't know where the file ends thanks to the guys that made the cgroups sysfs files have no size 02:19 < kurahaupo_> lektronx: regexes don't do sorting 02:20 < inthl> I have a friend's raspberrypi via SSH in my terminal. he has an issue I would know how to fix, but I am not entirely sure. There is 1 system on a SD card, the one that is currently up and I am connected to. His intention was to clone the SD card to another to get a copy of the installation. Now booting fails due to different UIDs - he can't fix it as easy, because of some hardware bug the keyboard is not working on init level1 where the 02:20 < inthl> report of the not-found UUID is shown 02:21 < kurahaupo_> lektronx: echo foo.[0-9]{,[0-9]} 02:21 < lektronx> kurahaupo_: touche... well what bash command could i use to sort the file listings? i know i've seen the info somewhere... 02:21 < inthl> now accessing that copied SD card ..how would I know to what the UUID is to be changed to in /etc/fstab ..? 02:21 < inthl> or as an alternative: is there a way to set it to device names or something? so that the copy could be cloned again, successfully booting without regard of the UUID ? 02:23 < phinxy> to tar a root fs would "tar --acls -cpf" be keeping all ownership and permissions? I guess -x makes sense to add. What about -a? 02:24 < mawk> I went for your suggestion ayecee , it makes very simple code maybe at the expense of some theoretical asymptotic complexity being higher 02:24 < mawk> my files are small anyway 02:25 < mawk> using 4 KiB on the stack is too much ? 02:25 < mawk> I forgot how much we have for our stacks normally 02:25 < phinxy> nvm -a is an option for cp, not tar. and -x as well(?) 02:25 < mawk> if it's a clone the uuid doesn't change I think inthl 02:25 < mawk> you maybe want to generate a new one 02:26 < dysfigured> any suggestions for decent routers i can install a regular ol linux distro on? 02:26 < mawk> why would you want that dysfigured ? 02:26 < mawk> instead of a regular small linux computer 02:26 < inthl> mawk, I guess it did not change. the issue is however that there comes up a initrd selection, from grub or so 02:27 < inthl> and once selected, the boot fails with some /dev/..uid not found 02:27 < mawk> strange 02:27 < inthl> the card is cloned properly I believe 02:27 < mawk> check the uuid with sudo blkid then 02:27 < gzuh> i don't think there's any "regular ol linux distro" you can install on a router. unless you get an x86 one or something 02:27 < mawk> compare with the one in /etc/fstab 02:28 < inthl> check where? 02:28 < inthl> on the running system or the cloned disk? 02:28 < dysfigured> mawk: well wouldn't a router have the specialized hardware a router would have, like antennae and..idk.. 02:28 < mawk> regular computers also have wifi antennaes and ethernet ports 02:29 < gzuh> dysfigured: what are you actually trying to do? there's projects like openwrt, and others i can't remember that are intentionally built for routers 02:29 < mawk> ddwrt, tomato thing 02:29 < lupine> regular ordinary debian is more than sufficient 02:29 < lektronx> karlhaworth[m]: thanks 02:29 < lupine> and it will install on most commodity router hardware, including ARM and MIPS boards 02:29 < dysfigured> really i just want a new router, and i wanted to put something like ddwrt on it 02:29 < lupine> (but I've got an APU2, for reasons) 02:29 < FutureTense> I installed a fresh copy of debian and then installed PiHole which set a static IP. Indeed I can SSH into that IP address again. 02:30 < dysfigured> i don' even know what ddwrt/openwrt would do for me, and i want to find ou 02:30 < lupine> mostly it's a relatively shiny web UI and a pretty rubbish packaging systen 02:30 < lupine> system* 02:30 < gzuh> dysfigured: when i got mine i just searched "openwrt" in amazon and got a cheapie travel router 02:30 < FutureTense> However when I do an ifconfig, it shows a different ip address. interfaces.d is alsho showing it is using DHCP 02:30 < FutureTense> what is going on? 02:30 < lupine> they also have prebuilt images for a range of surprising hardware 02:32 < cmj> "do you use pc or linux?" 02:34 < cmj> i swear this was just on moz://a.org 02:34 < hanetzer> one of these things is not like the other :P 02:35 < cmj> i will find it after dinner 02:35 < cmj> pc is windows? 02:35 < cmj> so weird 02:35 * cmj shakes fist at ibm 02:37 < mawk> on a raspberry pi FutureTense ? 02:37 < mawk> so you set a static ip for the raspberry pi, you ssh onto that static ip, and inside the ssh session you ran ifconfig or ip address and you saw a different address 02:37 < mawk> correct ? 02:38 < FutureTense> mawk, its a Rock64 02:38 < FutureTense> close to a pi, but not 02:39 < FutureTense> but otherwise, you have it corr 02:40 < mawk> paste output of ip address 02:42 < FutureTense> of the ifconfig? 02:43 < Sitri> No, of `ip addr show` 02:43 < Sitri> Oh wait misunderstood, ignore me 02:43 < phinxy> I moved a directory which had a 'bind' mount and then recreated the original directory which should have the 'bind' mount. However, mount -a wont make the files appear in the new directory. 02:47 < krzee> aww i missed Tj- 02:47 < krzee> sucks for me! 02:47 < krzee> but i knew that was the risk i took when work told me i had to go home lol 02:48 < dakd> is there a way to tell a program where to look for libraries? I have a 32 bit program that loads libraries from /usr/lib64 02:49 < phinxy> FutureTense• A couple released back on Ayufan's Debian for Rock64 I noticed there was two DHCP clients running. dhcpcd and dhclient. I ran 'apt purge dhcpcd' and then changed '/etc/NetworkManager/NetworkManager.conf' to have 'dhcp=dhclient' in it. IDK if its related or even if the latest versions have the two client running. 02:49 < phinxy> check htop for the two processes 02:50 < krzee> dakd: not really but you could use chroot and put libs in it's new root for it... thats all i got :D 02:50 < dakd> ouch 02:50 < krzee> unless maybe somebody else has a better way for you 02:51 < dakd> I am about to go back to 32 bit linux 02:52 < BlueProtoman> Does calling exec(2) (or one of its variations) free memory allocated by the calling process? 02:56 < phinxy> holy smokes, removing a bound directory removes the other link as well? 02:56 < lupine> it can, bind mounts are surprisingly configurable 02:57 < lupine> BlueProtoman: yup. most things are destroyed, including memory allocations. give it a go 02:57 < Sitri> BlueProtoman: read the man page 02:57 < Sitri> It's pretty thorough in what it lists 02:57 < FutureTense> arrrgh! somehow my dns is all messed up. I cant ping out 02:57 < FutureTense> can someone help me at least temprarily get this working? I'm not even sure which file to modify 02:59 < FutureTense> the interfaces file says it will be overrwitten and for me not to modify it 02:59 < AndroidKitKat> so I was working on my lockscreen 02:59 < AndroidKitKat> with this 02:59 < AndroidKitKat> https://www.reddit.com/r/i3wm/comments/5g86f1/suspend_on_lid_close/ 02:59 < AndroidKitKat> more specifically this: https://www.reddit.com/r/i3wm/comments/5g86f1/suspend_on_lid_close/daqel2p/ 03:00 < AndroidKitKat> and it was somewhat working 03:00 < AndroidKitKat> now it isnt 03:00 < AndroidKitKat> im not sure exactly what to do with the systemd conf file the poster suggested 03:00 < AndroidKitKat> what to name it, permissions, etc 03:00 < FutureTense> how can I find out where my DNS servers are set? 03:01 < AndroidKitKat> systemd is pretty weird imo 03:02 < BlueProtoman> Sitri: Didn't see it in the man page, especially since some variations of exec have their own man pages 03:02 < AndroidKitKat> does anyone know quite what to do? 03:02 < FutureTense> When I edit /etc/resolv.conf it tells me the changes will be overwritten 03:02 < BlueProtoman> lupine: Thank you 03:06 < AndroidKitKat> does anyone have any idea at all? 03:06 < AndroidKitKat> I'm not much of a systemd poweruser 03:07 < Psi-Jack> AndroidKitKat: WHat are you trying to actually do? 03:07 < AndroidKitKat> have i3lock run when i shut my laptop lid 03:07 < AndroidKitKat> or even better a bash script 03:08 < Psi-Jack> xautolock 03:08 < AndroidKitKat> that isnt available for my distro 03:08 < AndroidKitKat> oddly enough 03:08 < Psi-Jack> What distro? 03:08 < AndroidKitKat> Solus 03:08 < AndroidKitKat> solus-project.com 03:08 < Psi-Jack> Ah. Good luck with that. 03:09 < AndroidKitKat> it was working 03:09 < collins> off-topic. I'm burned out of developing things, going to flip burgers or something for a while so looking for something worth half my paycheck - well worth it. Is it wrong to use the phrase "unstimulating job" in my PM? 03:09 < AndroidKitKat> then i rebooted 03:09 < threexc> question for everyone in general 03:09 < Psi-Jack> collins: Very off-topic. 03:09 < threexc> Where is Linux actually used as a primary desktop OS in industry? 03:09 < Psi-Jack> threexc: My home. 03:10 < Psi-Jack> My work. 03:10 < threexc> It seems like most large companies prefer Windows 03:10 < meyou> developers mostly 03:10 < FutureTense> Is there a command like dpkg-reconfigure that will let you setup your ethernet as static? 03:11 < meyou> FutureTense, ubuntu? 03:11 < FutureTense> In other words, is there something that will do this that doesnt require me editing files? 03:11 < FutureTense> yes 03:11 < threexc> So what are some places I could work (in Canada) that might let me choose Linux, or provide it by default? 03:11 < meyou> FutureTense, nmcli 03:12 < meyou> https://help.ubuntu.com/community/NetworkManager#Using_NetworkManager_on_the_command_line 03:12 < Psi-Jack> threexc: Many places. 03:13 < FutureTense> is nmcli installed via sudo apt-get install network-manager 03:16 < meyou> it's usually included with a base install 03:16 < meyou> are you on an old ubuntu release? 03:16 < FutureTense> how do I tell what I have? 03:16 < FutureTense> I'm just a hacker with linux 03:17 < alexey-nemovff> AndroidKitKat: systemd suxks 03:17 < FutureTense> No, I'm running 16.04.4 03:17 < AndroidKitKat> alexey-nemovff: ik 03:17 < FutureTense> I just installed nmcli 03:17 < AndroidKitKat> but its in my distro 03:18 < FutureTense> but it tells me NetworkManager is not running 03:18 < AndroidKitKat> i just want an event to happen on suspend 03:18 < AndroidKitKat> is that so fucking hard 03:18 < Psi-Jack> AndroidKitKat: Kindly mind the language. 03:18 < AndroidKitKat> sorry 03:19 < FutureTense> ah sudo service network-manager start helped 03:20 < AndroidKitKat> im just confused as to why my system isnt suspending after i tell it to 03:20 < CodeBug> hey room need help with CentOS any good linux people in here :) 03:21 < FutureTense> ok, I've got nmcli running 03:21 < FutureTense> So how do I use it to set a static ip? 03:21 < xamithan> use the help or man page 03:22 < alexey-nemovff> AndroidKitKat: then your distro suxks while having the monster :v 03:22 < FutureTense> help is extremly sparse 03:22 < FutureTense> no man either 03:22 < AndroidKitKat> alexey-nemovff: solus is blazing fast 03:23 < AndroidKitKat> its like celtic magic makes this thing so fast 03:23 < xamithan> You sure? nmcli has a decent help function 03:23 < alexey-nemovff> AndroidKitKat: does Solus have it ? 03:24 < xamithan> "nmcli con add help" 03:24 < AndroidKitKat> have systemd? 03:24 < AndroidKitKat> sadly 03:24 < AndroidKitKat> the only distros w/o it are like void that i know of 03:24 < AndroidKitKat> maybe alpine 03:25 < alexey-nemovff> AndroidKitKat: what is your issue ? 03:25 < AndroidKitKat> im trying to have my laptop lock on suspend 03:26 < AndroidKitKat> following this reddit post 03:26 < AndroidKitKat> https://www.reddit.com/r/i3wm/comments/5g86f1/suspend_on_lid_close/daqel2p/ 03:27 < AndroidKitKat> the expected behavior is laptop suspends and locks 03:27 < AndroidKitKat> actual behavior, shutting and opening it just resumes where i was 03:29 < alexey-nemovff> I've a workaround but I'm in Void 03:29 < alexey-nemovff> and locking with xlock 03:29 < alexey-nemovff> AndroidKitKat: ^ 03:29 < AndroidKitKat> xlock isnt in my repos 03:30 < alexey-nemovff> :( 03:30 < AndroidKitKat> one of the downsides of solus 03:30 < alexey-nemovff> i see 03:31 < AndroidKitKat> so im trying to get systemd to do it 03:31 < AndroidKitKat> i just have no idea how to have that config file do what its supposed to 03:33 < FutureTense> ive no fucking idea what to do with nmcli 03:34 < FutureTense> I can see connections. No idea what to do with it. https://bpaste.net/show/552287219258 03:39 < Psi-Jack> FutureTense: Kindly mind the language., 03:40 < Psi-Jack> nmtui is easier to work with. 03:40 < Psi-Jack> AndroidKitKat: systemd doesn't run X apps, nor should it. 03:40 < Psi-Jack> That direction is incorrect. Hence why I suggested an x-based tool, xautolock, to do what you were asking, which is the correct way. 03:41 < AndroidKitKat> xautolock was it? 03:42 < Psi-Jack> I just said it, again, literally. 03:42 < AndroidKitKat> mb 03:42 < AndroidKitKat> trying to find a way to do it in solus 03:42 < Psi-Jack> Solus I don't support, it's not a very well maintained distro as a distro, and pretty much managed by one person. 03:43 < AndroidKitKat> yeah, ive been thinking of switching, but i dont know what to 03:43 < AndroidKitKat> maybe arch 03:43 < AndroidKitKat> just for support reasons 03:43 < AndroidKitKat> when you're the only person in the room running an obsure distro 03:43 < AndroidKitKat> it gets hard to help 03:44 < FutureTense> Psi-Jack, so I've opened nmtui 03:44 < Psi-Jack> Arch != support. Not recommended for newbies either. 03:44 < AndroidKitKat> not plain arch 03:44 < AndroidKitKat> like manjaro 03:44 < FutureTense> and I have two connections, ifupdown (etho) and etho 03:44 < AndroidKitKat> maybe debian testing 03:44 < Psi-Jack> Manjaro and other arch forks suck so much. 03:45 < AndroidKitKat> I used Antegros before 03:45 < AndroidKitKat> it sucked pretty bad 03:45 < FutureTense> It wont let me save the changes for ifupdown etho 03:46 < FutureTense> it says it is in read only 03:47 < Psi-Jack> Antergos isn't bad, it's mostly a GUI installer for arch, with some packages selectively in their own repo that isn't managed by Arch TUs. 03:47 < AndroidKitKat> but thoughts on it vs Manjaro? 03:47 < CodeBug> ??? 03:48 < Psi-Jack> I'm still not recommending Arch (or forks of), to people. Period. 03:48 < Psi-Jack> I use Arch myself, but I've been using Linux for ~24 years or so. Despite using it, I still won't recommend it to others. :) 03:49 < AndroidKitKat> I've been using Linux for about 8 months 03:49 < AndroidKitKat> started with CentOS, then Antergros 03:49 < FutureTense> Please please please. What do I need to do with nmtui to get a static ip? I've set it up, but when I reboot I'm still getting a DHCP address 03:49 < AndroidKitKat> and ive been using solus for 5 months 03:49 < Psi-Jack> FutureTense: Use your eyes. 03:50 < Psi-Jack> nmtui is sooooo easy to use. 03:50 < FutureTense> If it's so easy... why am I still getting a DHCP address? 03:50 < FutureTense> I edit a connection 03:51 < cyberbob> hi all, I'm planning to test ldap authentication for an inhouse developed app but we don't have any ldap here to test with. 03:51 < FutureTense> cant save the connection because it's read only 03:51 < Psi-Jack> FutureTense: Run nmtui as root. 03:51 < cyberbob> Can you champs please suggest some ldap really easy to setup and easy to manage, preferrably with some ui (as mostly QA/Dev teams will be managing that) 03:51 < Psi-Jack> QA/Dev managing LDAP? That's ASKING for trouble. 03:52 < cyberbob> Psi-Jack: agreed, but the environment is not that big and they don't have that many personals so we will be creating some doc to deploy and manage ldap so that they can take care of that . . 03:52 < cyberbob> So searching for something really/relatively easy 03:52 < Psi-Jack> Good luck with that. :p 03:53 < Psi-Jack> When you get QA and/or Devs to manage LDAP, you get bad LDAP. Pure and simple. 03:53 < Psi-Jack> There is no "really/relatively simple" LDAP. 03:53 < Psi-Jack> LDAP is LDAP. 03:53 < cyberbob> Psi-Jack: again agreed but believe me in current setup that is the best option for us. 03:54 < cyberbob> yeah but I was looking into Fedora directory 389 (I believe that is the name) . . .moved to some third party SSO and have not touched ldap thingy for quite a few years now 03:54 < littlepython> how can I know if a file system is read only? 03:55 < Psi-Jack> I don't believe you. Experience above "just do whatever you want" is so much better. 03:56 < cyberbob> Psi-Jack: :) 03:56 < Psi-Jack> cyberbob: FreeIPA (full 389-DS stack), 389-DS, or OpenLDAP. That's your primary choices. 04:00 < FutureTense> @Psi-Jack, I had to edit /etc/NetworkManager/NetworkManager.conf 04:00 < FutureTense> and changed "managed" to true 04:00 < FutureTense> then the connections appeared 04:00 < FutureTense> but still had to sudo 04:03 < Psi-Jack> FutureTense: This isn't Twitter, Slack, Discord etc. This is IRC. @nick notation is unusual and most irc clients don't do acting for it. 04:03 < FutureTense> Forget. I'm tired 04:06 < Psi-Jack> Man. My brain is turning into swiss cheese watching this TV show, Legion. heh 04:07 < FutureTense> Psi-Jack, nmtui is supposed to let you set DNS servers, correct? 04:07 < superkuh> Legion is a show without plot of coherence. It's only purpose is to allow them to create the most surreal looking scenes they can. 04:07 < superkuh> s/of/or/ 04:11 < AndroidKitKat> Psi-Jack: i decided to go with Debian Testing 04:11 < AndroidKitKat> its rolling enough where i can get my fix, but stable enough so i dont screw evertyhing up 04:11 < Psi-Jack> Good luck with that.... Expect unstable. 04:11 < FutureTense> this should not be so hard 04:16 < Psi-Jack> FutureTense: What distro are you on? 04:18 < FutureTense> Psi-Jack, 16.04 xenial 04:18 < FutureTense> this is on a odroid 04:18 < Psi-Jack> Ahh. 04:22 < mikhael_k33hl> Anyone using mailx? Does it need your host to be publicly available? can't seem to send an email using freebsd-mailx on an Ubuntu 16.04 workstation. 04:23 < Psi-Jack> No, you just need an MTA properly setup. 04:34 < garylabronz> hello everyone, so ima be googlin' but im not sure what to search for. i want to use the laptop A to share its wifi via ethernet to laptop B 04:34 < garylabronz> what do i google, as i need laptop A to do dhcp as well 04:34 < Psi-Jack> Why? 04:34 < Psi-Jack> A WAP is better suited for... Well, WiFi. 04:34 < garylabronz> me why? cos laptop b doesnt have drivers for its wifi until after ive installed em 04:35 < garylabronz> and i want to for funnsies 04:35 < garylabronz> am i "bridging" the networks? 04:35 < Psi-Jack> Replace the WiFi mPCIe underneath the keyboard with an Intel or Atheros. Simple. :) 04:35 < Psi-Jack> What WNIC do the laptops have? 04:35 < superkuh> Join the masquerade. 04:36 < garylabronz> yeah so anyways i apprecate the other ideas, but ya see its not the laptop in question its the OS. cos i put freeBSD on it 04:36 < superkuh> Keywords like "internet connection sharing" and "masquerade" will point you in the right direction, assuming you want more than just the local LAN shared. 04:36 < Psi-Jack> We don't support BSD here. 04:37 < garylabronz> luckily im not asking for help on that side ;) just how to make linux share its internet 04:37 < garylabronz> superkuh: thanks boss 04:38 < prussian> multiplexing your radio with one of the virtual ones being a base station and network bridging 04:40 < Abbott> `find ./ -name * -exec chown user:user {} \:` gives me "find: paths must precede expression: include" 04:40 < Abbott> what am I doing wrong? 04:41 < Abbott> oh I need to wrap the * in quotes 04:41 < Abbott> sorry lol I'll just see myself out 04:45 < prussian> dfor that you could do + instead of \; 04:45 < prussian> fyi 04:45 < prussian> chown -- user {} + 04:46 < prussian> i mean crap whatever chown user:grp -- {} + 06:03 < alexey-nemovff> I've never seen this channel this quiet.. :v 06:05 < saltystew> gotta think of something to talk about then 06:05 < alexey-nemovff> s/I've/I'd 06:06 < alexey-nemovff> saltystew: no rush xD 06:08 < G3nka1> I wanted to kill all java instances running, now I do ps aux | grep "java" and check each PID and do kill -9 on it. Is there a better way? 06:09 < [R]> killall -9 java 06:10 < G3nka1> thanks [R] can I kill java running only from my project? like not kill other java instances 06:10 < [R]> you have to kill it by pid 06:12 < G3nka1> so I should store all PIDS and kill them thats the only way 06:14 < [R]> well if you dont want to use ps... 06:14 < G3nka1> no it need not be just with ps I just want to stop all the java instances spawned from my project after a time [r] 06:15 < [R]> well you could use process groups 06:15 < [R]> and kill the whole thign 06:18 < G3nka1> how do you know that one java instance running is of one project and another o something else? 06:19 < [R]> what is a "project" 06:20 < G3nka1> [R] a project is the folder from which where I do java -jar .... 06:21 < [R]> that means nothing to linux 06:23 < G3nka1> right 06:23 < G3nka1> but I wanted to kill all these java servers started from this folder only 06:27 < saltystew> alexey-nemovff, I'm too busy watching bsg for the first time 06:27 < ghulan> bsg... 06:27 < saltystew> battlestar galactica? 06:27 < alexey-nemovff> bag? 06:27 < alexey-nemovff> bsg? 06:28 < ghulan> *save* for later. 06:28 < saltystew> it's awesome so far 06:35 < G3nka1> $ ls 06:35 < G3nka1> client.sh* common.sh* server.sh* start-all.sh* stop-all.sh 06:35 < G3nka1> does these * mean they're symbolic links? 06:35 < [R]> use ls -l 06:35 < alexey-nemovff> executable 06:36 < alexey-nemovff> if they were symlinks there'll be a > followed by the actual target 06:36 < alexey-nemovff> G3nka1: ^ 06:36 < G3nka1> how do I make my .sh file an executable? 06:36 < G3nka1> ah I see 06:37 < [R]> G3nka1: chmod 06:37 < alexey-nemovff> chmod +x 06:37 < G3nka1> did that still its not executable 06:38 < G3nka1> ls 06:38 < G3nka1> client.sh* common.sh* server.sh* start-all.sh* stop-all.sh 06:38 < G3nka1> chmod +x stop-all.sh 06:38 < Psi-Jack> ls -l 06:38 < Psi-Jack> long format shows permissions 06:38 < G3nka1> -rw-r--r-- 1 Administrator 197121 7 Apr 10 10:05 stop-all.sh weird I just did +x and even 777 06:39 < G3nka1> still its like this 06:39 < Psi-Jack> What filesystem is that on? 06:39 < Psi-Jack> Looks like.... NTFS? 06:39 < G3nka1> Psi-Jack, I am windows with git bash 06:39 < [R]> rofl! 06:39 < Psi-Jack> We don't support Windows here. 06:40 < sauvin> I, for one, would be *unable* to support Windows or "git bash". 06:51 < alexey-nemovff> Winblows sucks 06:51 < ghulan> For the least, Win 10 will support Win 10+ API. 06:52 < ghulan> Multi env in LAN is a good thing. 06:52 < Psi-Jack> Not really relevant to Linux though. 06:52 < ghulan> Yeah... 08:05 < ShapeShifter499> hi 08:05 < Roserin> hi 08:06 < Psi-Jack> konnichiwa 08:07 < ShapeShifter499> I'm starting to end up with a sizable number of ARM based headless Linux devices running various up to date Linux distros. So far Arch Linux and Debian, what do you guys suggest I do to manage the login for all of these devices? 08:07 < Psi-Jack> Ldap 08:07 < Roserin> screen+ssh? 08:08 < ShapeShifter499> Psi-Jack: こんばんは 08:09 < ShapeShifter499> Psi-Jack: I've heard of ldap but I have never set up anything like it 08:09 < ShapeShifter499> Roserin: I'd have to manage a ton of passwords in this case 08:12 < ShapeShifter499> Psi-Jack: is there good place to start with LDAP? 08:15 * iflema Roku 08:16 * iflema singer 08:16 < Dagmar> well, there's always YP or Kerberos, but some people find it simpler to just start cutting their arms with hobby knives 08:18 < Dagmar> If you're into Lutheranism, you can always try AD 08:19 < ShapeShifter499> Dagmar: what? 08:19 < Dagmar> Active Directory 08:19 < Dagmar> Although personally I think people who deploy AD to manage Linux login domains should probably be required to register their home address with a government agency and not live near schools 08:19 < ShapeShifter499> Dagmar: I didn't get the joke sorry lol 08:20 < Dagmar> Lutheran orthodoxy involves moritifcation of the flesh 08:20 < Dagmar> i.e., self-flagellation 08:21 < ghulan> It's a charge for the service. Though there's many where that came from. 08:23 < ShapeShifter499> Dagmar: jokes aside. I'm just looking for the best way to manage a ton of linux logins without going mad trying to figure out which password went where 08:23 < ShapeShifter499> I was considering just setting up a bunch of secured ssh keys and password protect them with my password manager 08:27 < jimm> I wrote this whole flatfile db thing (I called it the roster project) for keeping track of students in classes that need particular server logins 08:27 < jimm> it kept track of hundreds of logins 08:30 < ShapeShifter499> I wonder if there is a "password manager" like app just for logins 08:30 < ayecee> yes, there is 08:30 < ShapeShifter499> ooo 08:30 < ShapeShifter499> I could really use that 08:30 < ayecee> keepass is the one that comes to mind 08:30 < ShapeShifter499> no wait 08:31 < ShapeShifter499> I was imagining a terminal program or CLI based program where I could input my secure password to decrypt things then select which server I wanted to connect to 08:32 < hexnewbie> Depending on what logins mean, the answer may be SSH keys, SSH agent and/or DM 08:32 < ayecee> kind of like hwat keepass does 08:32 < ShapeShifter499> ayecee: I already have keepass 08:34 < ShapeShifter499> hexnewbie: DM? 08:35 < ShapeShifter499> I have never used any LDAP or similar methods, I'm still pretty new to logins 08:35 < ayecee> you are still young enough to learn 09:04 < hehehe> hey hey 09:04 < hehehe> who is ere? :P 09:04 < Roserin> 09:05 < hehehe> lol 09:09 < sauvin> 09:09 < EugenA> looking for lightweight tool to draw lines / rectangles on screenshots 09:10 < sseeley> apple pencil 09:10 * sseeley ducks 09:23 < ice9> jdk10-openjdk is installed however and app is saying: Java JDK not found! Install in standard location or set java_paths! what paths I should set?? 09:24 < vlt> EugenA: Something like Inkscape? 10:11 < sauvin> Zz. 10:18 < nai> shh, people are trying to sleep here 10:19 < webchat256> Hi all i need some help with routing i have a box with 3 nics, the first is connected to the internet using dhcp other two are static ips with no gateway in different subnets ive setup ip forwarding and masquerading the problem is i have no internet when i bring up the third nic when using only two one connected to internet and other to private network all works fine, any help would be greatly appriciated 10:21 < Cybertinus> hello everybody 10:22 < Cybertinus> I'm trying to configure my local ssh configuration to apply a certain configuration to all hosts in a domain (so: hosts *.domain.tld). But for 1 host in this domain, I don't want to to apply the *.domain.tld options. How can I do this? 10:22 < nai> wew 10:24 < Cybertinus> right, IRC did it's magic trick again 10:24 < Cybertinus> found it 10:24 < Cybertinus> Host *.domain.tld !specifichost.domain.tld 10:24 < Cybertinus> in ~/.ssh/ssh_config 10:53 < ice9> what should be the standard JAVA_HOME path? 11:02 < revel> ice9: Probably depends on your distro. It's /usr/lib/jvm/oracle-jdk-bin-1.8/ on my machine. 11:03 < revel> And on your JVM variant and its version, of course. 11:16 < VjdfMQ> Hey all 11:16 < VjdfMQ> Why htop gives more than 100% percents ? https://i.imgur.com/CwkNxZk.png 11:16 < Hooloovo0> do you have more than 1 cpu? 11:16 < the_drow> VjdfMQ, because some operations can be parallelized 11:16 < Hooloovo0> most computers do nowadays 11:16 < the_drow> Even on the same CPU 11:16 < the_drow> See SIMD 11:17 < Hooloovo0> I've had up to 2400% 11:17 < the_drow> That's a lot 11:17 < Emil> Hooloovo0: on a singe cpu? 11:17 < Hooloovo0> yeah, but I don't think simd is counted in thread accounting 11:17 < Hooloovo0> no, on 24 cpus 11:17 < Emil> hah 11:17 < VjdfMQ> Hooloovo0: Yes, you mean it can show 400% if ther're 4 cores ? 11:17 < Emil> Hooloovo0: seti@home? :D 11:17 < Hooloovo0> yeah 11:18 < Hooloovo0> boinc has done that, also several other programs 11:18 < Hooloovo0> most recent I think was a conway's game of life search 11:18 < Emil> ah it was boinc 11:19 < Hooloovo0> most recent when I maxed all cpus was a gentoo upgrade, but that doesn't spawn one process with many threads, just a lot of threads 11:19 < Emil> :D 11:19 < Emil> install gentoo 11:19 < the_drow> I have a LXC container with a iptables rule that limits the connection to specific IP addresses. I'm writing a test suite using serverspec that has a test that checks connectivity from those IP addresses. I'm trying to find a way to add my computer as a node in lxcbr0 with a specific IP without creating another LXC container. 11:20 < the_drow> I tried iptables -t nat -A POSTROUTING -p tcp -j SNAT --to-source 10.0.3.3 --destination 10.0.3.133 on my computer but packets are not routed back to my IP. 11:20 < the_drow> Packets are sent with 10.0.3.3 though so that's half the job 11:20 < the_drow> Maybe there is another solution that does not involve iptables but creating an actual route 11:21 < the_drow> Any ideas? 11:22 < edd_lc> the_drow: How about do the same on the other end? POSTROUTING to your actual IP when destination is 10.0.3.3. But, to be sincere, I don't know what an LXC container is. Maybe it's not possible 11:22 < the_drow> Just a wrapper around namespaces and stuff 11:23 < the_drow> Used to be the backend for Docker. 11:24 < the_drow> edd_lc, Note that 10.0.3.3 does not really exist 11:24 < the_drow> I'm trying to fake it in a way that will route this IP to my computer 11:24 < TJ-> the_drow: still at it? have you pasted the network config? bridges, interfaces, routing table, etc? 11:25 < promach_> hi, I am trying to commit a local change to my github repo, but git push told me "Everything up-to-date". Why ? 11:26 < the_drow> TJ-, Yes, I had to quit yesterday 11:26 < the_drow> TJ-, I haven't pasted them yet 11:27 < the_drow> TJ-, ip a output: http://dpaste.com/1ZT783H 11:27 < the_drow> TJ-, brctl show output http://dpaste.com/34E42H4 11:28 < the_drow> promach_, not the right channel. but you need to set your branch to upstream 11:28 < the_drow> git branch --set-upstream-to=origin/branchname branchname 11:28 < the_drow> promach_, go to #git 11:29 < the_drow> TJ-, routing table http://dpaste.com/0N5K1TG 11:30 < promach_> phung@UbuntuHW15:~/Documents/riffa$ git branch --set-upstream-to=https://github.com/promach/riffa/tree/development development 11:30 < promach_> fatal: branch 'development' does not exist 11:31 < the_drow> promach_, not the right command 11:31 < promach_> the_drow: sorry for posting in wrong channel, but where would I ask this ? 11:31 < the_drow> at #git 11:31 < the_drow> TJ-, So any ideas? 11:35 < TJ-> the_drow: remind me of the iptables rule you are setting, and the destination address 11:35 < the_drow> iptables -t nat -A POSTROUTING -p tcp -j SNAT --to-source 10.0.3.3 --destination 11:36 < TJ-> the_drow: and what is the container IP? 11:36 < the_drow> Whatever I want it to be 11:37 < TJ-> the_drow: what is it /right now/ though, with the container running. 11:38 < TJ-> the_drow: the container is on the local host, correct? so whatever IP sub-net it is on should be in the routing table 11:38 < the_drow> TJ-, yes localhost 11:39 < the_drow> TJ-, 10.0.3.133 11:39 < the_drow> Why does it matter? 11:39 < the_drow> It's on the same subnet 11:39 < TJ-> does the container respond to 'ping 10.0.3.133' done on the host? 11:40 < TJ-> the_drow: the routing table you showed does not show any 10.x subnet, therefore those packets will be leaving on the default route. You appear to have 2 default routes so I assume 1 has a lower metric than the other, but either way, those packets are leaving the host and going out over WiFi or Wire 11:41 < the_drow> So how come two LXC containers can communicate with each other through lxcbr0? 11:44 < the_drow> TJ-, for some reason lxcbr0 is not listed in the routes 11:45 < TJ-> the_drow: because it's a bridge so they appear on a single segment 11:45 < the_drow> TJ-, no because ICMP is blocked 11:45 < the_drow> Everything but SSH is blocked right now 11:46 < the_drow> The rules work in production but I have to find a way to test them in CI 11:47 < TJ-> the_drow: there is no way to route to the container from the host since there is no entry in the routing table for the container subnet, which requires lxcbr0 to have an IP address 11:47 < the_drow> lxcbr0 does have an IP address 11:47 < TJ-> the_drow: no, it does not. End of http://dpaste.com/1ZT783H 11:47 < the_drow> 10.0.3.1 11:49 < the_drow> TJ-, I restarted lxc-net 11:49 < the_drow> Now it does 11:50 < the_drow> TJ-, http://dpaste.com/115MSDP 11:50 < the_drow> new routing table http://dpaste.com/19RG2WT 11:53 < the_drow> Btw iptables nat table has -A POSTROUTING -s 10.0.3.0/24 ! -d 10.0.3.0/24 -j MASQUERADE 11:53 < TJ-> the_drow: so what netfilter rules are in place right now ("iptables -S") 11:54 < Sharetel> Hi, can anyone please help me with a request - I need to fetch the type of files such as .docx, .pdf, .xlsx etc recursively along with their total size in MBs, GBs. Please help. 11:54 < Sharetel> "find . -type f -name "*.mp3" -print0 | du -hc --files0-from=- | tail -n 1" I tried this but didn't help 11:54 < the_drow> TJ-, http://dpaste.com/3Q67MPZ 11:55 < VjdfMQ> Could anyone suggest why Suprt key stops working after some time ? I mean, xev shows that it captures, but it doesn't call a start menu 11:55 < widp> How do I set default jack volume? 11:56 < the_drow> TJ-, iptables -t nat -S http://dpaste.com/3DZPG8W 11:56 < widp> It starts up at 633% for me. 11:56 < widp> leading to very distorted sound. I just end up closing jack unless I need it. 11:56 < the_drow> Sharetel, https://stackoverflow.com/questions/28900112/how-to-calculate-the-total-size-of-certain-files-only-recursive-in-linux 12:01 < the_drow> TJ-, So, any ideas? 12:01 < TJ-> the_drow: what is the actual real source address of the packets being spoofed? I suspect one of those othee MASQUERADE rules is matching first 12:02 < the_drow> oh 12:02 < the_drow> my ip 12:02 < the_drow> I'm wondering if it's localhost or my actual IP 12:02 < the_drow> I'll check 12:03 < TJ-> the order of the rules might be the issue - assuming for the moment the SNAT rule is correct. If the SNAT is after those others, the packet may already have been NATed and not hit your SNAT rule 12:05 < the_drow> TJ-, Yup, they come after 12:06 < the_drow> TJ-, I'm trying iptables -t nat -I POSTROUTING -p tcp -j SNAT --to-source 10.0.3.3 --destination 12:07 < Sharetel> the_drow, thank you but it shows output in huge numbers and not human readable as in MBs or GBs 12:07 < TJ-> the_drow: I'd insert some -j LOG rules in amongst them and check dmseg to get an idea of where the packets are going. you could create a custom chain for your requirement, do -j custom_chain then have a LOG rule followed by the actual SNAT rule in custom_chain 12:07 < the_drow> Sharetel, use your shell foo 12:07 < the_drow> s/foo/fu 12:08 < the_drow> TJ-, No luck even if the rule is before the MASQUERADE rule 12:09 < Sharetel> ok, thanks the_drow 12:09 < jimm> what's a shell foo?! 12:09 < the_drow> I meant shell-fu 12:10 < TJ-> Sharetel: " find path/to/dir -type f -name "*.mp3" -ls | awk '{sum += $7} END {print sum / 1024^2 "MB"}' " 12:10 < the_drow> It's a computer-martial-arts style that belongs to the neckbeard sysadmin philosophy 12:10 < TJ-> oh, he gone! 12:11 * the_drow has great timing but zero delivery capability 12:12 < the_drow> It sounded way better in Hebrew 12:13 < djph> the_drow: I doubt that. 12:13 < the_drow> LOL 12:14 < the_drow> TJ-, Would you mind telling me where to place my logs? 12:15 < the_drow> I'm not sure I understand your suggestion 12:15 < djph> it probably sounded leftways, instead of rightways. (I mean, I bet if you knew it, maybe not ... but I can barely follow along in EN) 12:16 < the_drow> hehe 12:40 < the_drow> TJ-, I can't for the life of me figure out how this works 12:43 < djph> the_drow: what logs? 12:43 < the_drow> iptables logs 12:43 < the_drow> I never used those 12:43 < the_drow> Not sure where to place them 12:48 < the_drow> This is the tcpdump output http://dpaste.com/0111KNC 12:48 < the_drow> Not that I'm running tests against two different containers 12:48 < the_drow> s/Not/Note 12:50 < the_drow> djph, I have a LXC container with a iptables rule that limits the connection to specific IP addresses. I'm writing a test suite using serverspec that has a test that checks connectivity from those IP addresses. I'm trying to find a way to add my computer as a node in lxcbr0 with a specific IP without creating another LXC container. I tried iptables -t nat -A POSTROUTING -p tcp -j SNAT --to-source 10.0.3.3 --destination 10.0.3.133 12:50 < the_drow> on my computer but packets are not routed back to my IP 12:50 < the_drow> I think I have a missing rule to route 10.0.3.3 back to my real IP but not sure 12:57 < djph> 10.0.3.3 is ... what? 13:02 < kurahaupo_> the_drow: what do those 2 addresses refer to? 13:03 < the_drow_mobile> Two lxc containers 13:06 < the_drow_mobile> djph, it's a spoofed IP address 13:09 < ychaouche> hello ##linux 13:10 < ychaouche> Is /var/log/syslog a file that would contain lines from the different facilities ? like, are lines duplicated in the facility file + syslog ? 13:10 < noodlepie> Hiya ychaouche 13:10 < BluesKaj> Hi folks 13:11 < noodlepie> hi BluesKaj 13:11 * ychaouche should I read it blue sky ? 13:11 < BluesKaj> hi noodlepie 13:11 < BluesKaj> ychaouche, if you wish 13:11 < ychaouche> Grant me three wishes 13:12 < BluesKaj> it's an obvious play on words 13:12 < ychaouche> you player you 13:14 < BluesKaj> i also like the blues and play drums in a blues/rock band 13:14 < BluesKaj> hence the nick 13:15 < the_drow_mobile> djph, still around? 13:16 < ychaouche> I told you you were a player 13:17 < rickardo1> How can the exactly same file show different output with file -i on different servers? styles.ff95522fb05496eab506.bundle.css: application/octet-stream; charset=binary and styles.ff95522fb05496eab506.bundle.css: text/plain; charset=us-ascii on the other? 13:20 < frostschutz> rickardo1, if the checksum is identical, you have different versions of file / magic 13:22 < frostschutz> file is a heuristic, the results are not reliable especially not across different distros / installs. you can even define your own signatures for file in your homedir 13:23 < beaky> hello. how do i tell my terminal to ping me 5 minutes from now 13:25 < kurahaupo_> beaky: sleep 300 ; printf \\a 13:25 < lukey_> the_drow: "ifconfig lxcbr0:1 up " 13:25 < beaky> wow thanks kurahaupo_ that works nicely 13:30 < beaky> what i didnt heear anything 13:34 < fendur> you mean you didn't test it with 1 second sleep? 13:36 < kurahaupo_> beaky: you need to enable "bell" in your terminal emulator 13:36 < beaky> oh 13:37 < kurahaupo_> After you've done that, be extra careful never to cat a binary file 13:42 < VjdfMQ> Could anyone suggest why Super key stops working after some time ? I mean, xev shows that it captures, but it doesn't call a start menu 13:42 < elichai2> hey 13:43 < elichai2> Anyone knows what's the difference between "Headset", "Handsfree" and "Audio Sink" in blueman? 13:48 < gebbione> hi folks, is there a way to run a command for a user like nginx that doesn't have a terminal? 13:48 < gebbione> i m trying to troubleshoot permission errors 13:49 < gebbione> sudo su nginx -c ls /var/www -> This account is currently not available. 13:50 < the_drow> lukey_, I'm gonna try that now 13:51 < kurahaupo> gebbione: "having a terminal" is an attribute of a process, not of a user 13:53 < kurahaupo> gebbione: sudo -u nginx ls … 13:54 < the_drow> lukey_, SIOCSIFFLAGS: Cannot assign requested address 13:55 < gebbione> kurahaupo, thank you 13:58 < the_drow> lukey_, Ok I need to use ifconfig before creating the iptables rule 14:02 < the_drow> lukey_, Can you explain why this works? 14:02 < the_drow> Also, how do I remove it? 14:03 < BCMM> elichai2: not sure about blueman specifically, but it looks like those refer to different bluetooth profiles 14:03 < elichai2> couldn't find online whats actually the diffrence 14:03 < BCMM> elichai2: a headset is bidirectional mono audio (i.e. microphone and single earpiece) 14:03 < BCMM> elichai2: handsfree is same as above, but with the ability to control the device (e.g. headset with an "answer" button) 14:04 < BCMM> elichai2: audio sink is unidirectional and can support stereo 14:05 < BCMM> elichai2: many devices can switch between profiles. i've got a device that acts as a handsfree for calls, but acts as a stereo audio sink with no mic when listening to music 14:06 < elichai2> yeah, I thought maybe one of these profiles recognize that mine support couple and will automatically switch 14:06 < elichai2> right now I'm manually switching between HSP and A!DP 14:06 < elichai2> *A2DP 14:07 < BCMM> elichai2: I think hsp = "Headset" and a2dp = "Audio Sink" 14:09 < djph> the_drow: sorry, some helldesk dweeb gave someone my number. err, if you're spoofing the IP, assuming the target machine answers, the answer will be sent to the wrong (or nonexistant) machine 14:09 < the_drow> djph, ifconfig lxcbr0:1 up '10.0.3.3' worked once I added it manually but now doesn't 14:10 < the_drow> Also, not sure what I typed yet... 14:10 < djph> the_drow: hmm ... 14:11 < djph> the_drow: bridge interface bridging the right container interface to teh right host interface? 14:11 < the_drow> Essentially I'm trying to make LXC "think" my computer is just another LXC container with the IP I provided 14:11 < the_drow> djph, yup 14:11 < ghulan> Does anyone know where I can get a cool background for my DE. 14:11 < ghulan> ? 14:12 < djph> ghulan: white text on a black background? 14:13 < ghulan> It's gotta be a picture, obviously. 14:13 < kurahaupo> ghulan: most chemistry labs have a cryo freezer 14:13 < JimBuntu> ghulan, look no farther than your webcam or phone cam, take a selfie. 14:13 < ghulan> I try... 14:14 < djph> take a screenshot of white text on a black background? 14:14 < the_drow> djph, the question is if I should use up or add. up provides me with 10.0.3.3/8 and add provides me with 10.0.3.3/0 14:14 < ghulan> the_drow, There is no /0. 14:14 < djph> i think you're doing it wrong if something's giving you /0 14:14 < the_drow> There is no sppon 14:15 < elichai2> BCMM: so nothing to automatically change : 14:15 < elichai2> :\ 14:15 < the_drow> djph, so up it is but it gives me /8 14:15 < the_drow> I want /24 right? 14:15 < BCMM> elichai2: well, under what circumstances? 14:15 < ghulan> Any # between /~8-24 is good. 14:15 < BCMM> elichai2: i mean if you have a viop program that lets you run an arbitrary command when you get a call, like a good voip program ought to... 14:15 < djph> the_drow: dunno, what's the existing netmask? 14:16 < the_drow> inet 10.0.3.3/8 brd 10.255.255.255 scope global lxcbr0:1 14:16 < elichai2> BCMM: I do video chat in the browser... 14:17 < the_drow> I keep getting SIOCSIFFLAGS: Cannot assign requested address even though the address gets added 14:17 < ghulan> Even if you did, you still need a tunnel from someone else. So, there's always got to be one host up for network. 14:20 < lukey_> the_drow: : lets you add multiple ip-adresses to the same network interface 14:20 < ghulan> And as far as it goes, the common RJ-45 will be needed for a host that's miles away. 14:21 < the_drow> lukey_, Really? You just made my day :) 14:21 < the_drow> ip address add 10.0.3.3/24 dev lxcbr0:1 does not produce any errors but ifconfig does 14:26 < lukey_> the_drow: Hmm is the address already assigned? 14:26 < the_drow> It wasn't 14:26 < the_drow> Anyway my test fails now because ssh requires a password and it shouldn't 14:27 < ghulan> Which kind of set up are you looking at? 14:28 < CtrlC> Is there anyway I can make proxychains to give a free pass to localhost requests? 14:28 < the_drow> ghulan, I'm trying to spoof my IP so that I can communicate with a LXC container as if I was another LXC container with some IP 14:29 < ghulan> Search "gateway." 14:30 < ghulan> It mostly boils down to that. Though you need an antenna with PoE to make it work. 14:30 < lukey_> the_drow: To or from the host? 14:30 < mawk> what's the equivalent of std::move for template arguments already ? 14:30 < lukey_> the_drow: SSH 14:30 < the_drow> lukey_, I'm establishing an ssh connection to the container from the host 14:32 < blackflag_bfp> @item(buffers):button1 14:32 < blackflag_bfp> whoops sry about that and good morning 14:34 < lukey_> the_drow: Hmm. You could also try "firejail --net=lxcbr0 --ip= bash" and do the stuff you want inside this shell 14:34 < the_drow> firefail? 14:34 < the_drow> lukey_, This is an automated test written with serverspec (rspec) 14:38 < lukey_> the_drow: You can also run your testsuite inside the firejail 14:38 < the_drow> what's a firejail? 14:39 < the_drow> oh another container technology? 14:40 < lukey_> the_drow: Not quite. But containerisation and Jailing share some stuff 14:41 < the_drow> lukey_, I can run a specific test command with firejail I guess. I'd rather narrow down the dependencies of the test suite but maybe if it makes things easier I'll do so 14:44 < the_drow> lukey_, You just made my day 14:44 < the_drow> Thank you 14:54 < flashdel> hi folks, i got a question: I would like to use nosuid on a nfs share i got and i would like to know how i can find out if i just need the remount command to remount it or if i need to do it "the hard way". Can somebody help? 14:54 < jhodrien> flashdel: Doesn't this come with an easy answer? 14:54 < jhodrien> Stick something suid on it, try it, find out. 14:55 < flashdel> jhodrien: i would like to find it out without testing it, but i cannot find something in the nfs man page 14:55 < jhodrien> Why not mount the share on another machine and test? 14:56 < jhodrien> Why not create a test share to test? 14:57 < flashdel> i am still in a learing phase and i want to read manpages the right way or find things out the right way. In this case, you are right, i could just test it, but i want to find it out by researching 14:58 < dgurney> testing is one method of researching something 14:58 < jhodrien> You'd expect it to work from the man page 14:58 < jhodrien> You'd believe it'll work after testing it. 15:00 < flashdel> i know your point and you are right, but i dont even expect something, because i just dont know it, because there is nothing written in the manpage 15:00 < ghulan> Quit the BS. This isn't BSD. 15:02 < P_B> O_o 15:03 < ghulan> Analytica mostly uses BSD anyway. I don't follow with that. 15:03 < mawk> chrome://flags/#enable-parallel-downloading 15:03 < mawk> nice feature 15:04 < BCMM> isn't parallel download basically "cheating"? exploiting contested servers that aim to give every connection equal bandwidth, to give your user a faster download at the expense of other users? 15:05 < djph> BCMM: suppose it could be that, sure. 15:08 < mawk> yeah BCMM , said like that 15:08 < mawk> but I'd expect the servers to throttle by IP, not by connection ? 15:08 < mawk> or maybe it would break public connections 15:08 < the_drow> lukey_, Do you know by any chance how to pass multiple arguments to firejail? 15:09 < the_drow> I mean to the program firejail is executing 15:12 < BCMM> last time this sort of thing was popular (late '90s, early '00s - spammy ad-filled "download accelerators for windows), download mirrors responded by limiting the total connections from a single ip 15:12 < BCMM> i guess that was easier than changing bandwidth allocation from per-connection to per-address? 15:13 < ghulan> BCMM, Power is knowledge. 15:13 < BCMM> if that happens again now, it's gonna suck even worse than it did then, because there's a lot more people on NAT 15:14 < djph> BCMM: yeah, CGN is awful -- but these days we have CDNs, as opposed to "my one webserver..." 15:17 < mawk> the_drow: firejail ...firejail options... -- program arg1 arg2 15:17 < mawk> this should work no ? 15:17 < BCMM> even worse is if mirrors *can't* take meaningful action against this, and then everybody has to do it to remain competative 15:18 < the_drow> mawk, well it does but only takes one argument 15:18 < the_drow> ssh 10.0.3.54 works 15:18 < mawk> that's strange the_drow 15:18 < BCMM> and when everybody is doing it it no longer has significant benefit, just a bunch of overhead for nothing 15:18 < mawk> you're sure you're not using quotes ? 15:18 < the_drow> The rest is not parsed it seems 15:18 < mawk> how do you test it ? using echo ? 15:19 < the_drow> mawk, Using the ssh command line 15:19 < the_drow> It seems to ignore -oConnectTimeout=50 and exit after a few seconds 15:19 < the_drow> maybe I'm missing something 15:19 < mawk> add my -- as I said 15:19 < the_drow> already did 15:20 < mawk> strange 15:20 < mawk> try quoting everything then 15:20 < mawk> firefail $OPTS 'ssh -o stuff=foobar' 15:29 < nng> hi, i have one linux server where i cant connect to another i have this issue: ssh_exchange_identification: read: Connection reset by peer 15:29 < nng> from this server i cant connect also to another 15:30 < nng> but from my notebook i can connet succefull to both 15:30 < nng> any sugestions? 15:31 < Psi-Jack> nng: Private keys the same or different? 15:32 < nng> public keys are add to authorized_keys 15:33 < Psi-Jack> You didn't answer the question. 15:35 < nng> different 15:35 < Psi-Jack> Permissions correct on the one that's failing? 15:36 < nng> i think so 15:36 < talx> does it happens only on ssh ? 15:37 < nng> i just try ssh 15:37 < talx> do you get a ping reply ? 15:37 < nng> ping its pass 15:38 < Psi-Jack> ls -ld ~/.ssh; ls -l ~/.ssh/ -- to paste site (not pastebin.com) 15:38 < talx> oh nvm I missed few lines u wrote 15:38 < Psi-Jack> "you" not "u" 15:38 < Psi-Jack> "nevermind" not "nvm" 15:39 < kurahaupo> Psi-Jack: ls -la includes . and .. which are both useful 15:39 < ne0crysis> nng: take a look at hosts_allow maybe 15:39 < Psi-Jack> Hmm, true. :) 15:39 < talx> hmm okay power per inch-jack 15:39 < talx> or I don't know 15:39 < Psi-Jack> talx: ... 15:39 < Psi-Jack> I guess you don't want help. :) 15:39 * kurahaupo wnds up Psi-Jack's grmmr ftsh 15:39 < talx> heh it's not that 15:40 < talx> help is always needed 15:40 < BluesKaj> talx, it's proper English here not testspeak 15:40 < kurahaupo> talx: pounds per 15:40 < kurahaupo> BluesKaj: textspeak? 15:41 < talx> it doesn't say perfect english in the topic 15:41 < BluesKaj> texting shortforms on mobiles , kurahaupo 15:41 < Psi-Jack> shtspk, sms-speak, text speak, however you identify it with.. Is not desired here and specifically mentioned in the channel rules. Not everyone here is native English, and we like all to understand. 15:41 < Psi-Jack> talx: Actually it does. In the topic's URL rules page. 15:42 < kurahaupo> talx: on the other hand, I concur with Psi-Jack that "U" and "ur" are especially grating 15:42 < nng> hosts.allow and .deny are empty 15:42 < talx> yea I'm not arguing or anything haven't seen that link tho 15:42 < talx> though* 15:42 < nng> https://pastebin.com/Whq15uzd 15:42 < talx> Yeah* 15:43 < junka> U2 15:43 < Psi-Jack> talx: Well, you're 100% responsible for reading the rules of the channel you're in. 15:44 < Psi-Jack> You come here for help, when corrected you get offensive. How do you think that helps you? ;) 15:44 < kurahaupo> junka: 🎶 and I still haven't found what I'm looking for… 15:45 < talx> how come I'm offensive heh 15:45 < junka> kurahaupo; im rooting for you 15:45 < Psi-Jack> Good luck. talx. 15:45 < kurahaupo> Psi-Jack: OTOH over-zealous enforcement can make bystanders nervous 15:46 < fendur> kurahaupo++ 15:48 < talx> Psi-Jack: thank you. sorry if you got offended by 'sms-talk' didn't mean to hurt you. english is not my native language so I don't use it in sms but I get the idea 15:49 < Psi-Jack> Eggselent. ;) Now I'm willing to help you again. So, I need the output of 'ls -l ~/.ssh/' on the client that's failing to connect. 15:49 < junka> can i have 2+4 GB of ram? 15:49 < dgurney> why not 15:49 < Psi-Jack> junka: ##hardware 15:49 < dgurney> there's nothing that prevents you from having 6GB of ram... 15:50 < junka> dgurney; cool :P 15:50 < junka> Psi-Jack; thanks 15:50 < kurahaupo> junka: 6GB in what? What does it currently have? 15:51 < Psi-Jack> I referred ##hardware because such a question leads to more questions, more questions which relate more to hardware specifics, and ##hardware is a channel more specific to that topic. :) 15:52 < nng> Psi-Jack: https://pastebin.com/psR7gGL3 15:52 < nng> Psi-Jack: from client 15:53 < Psi-Jack> Never use pastbin.com 15:53 < Psi-Jack> I mentioned that earlier. :) 15:53 < nng> why? 15:53 < nng> there no pass etc 15:53 < Psi-Jack> nng: Many reasons. They destroy pastes, many companies block the site. 15:53 < Psi-Jack> Ads, a big one, including malvertising. 15:54 < nng> i thinh the issue is on server where i try connent 15:55 < nng> becouse he can't connet to any servers 15:56 < kurahaupo> nng: try ssh -vv (and the other options you're already using). Note the output 15:57 < nng> OpenSSH_6.7p1 Debian-5+deb8u4, OpenSSL 1.0.1t 3 May 2016 15:57 < nng> debug1: Reading configuration data /etc/ssh/ssh_config 15:57 < nng> debug1: /etc/ssh/ssh_config line 19: Applying options for * 15:57 < nng> debug2: ssh_connect: needpriv 0 15:57 < nng> debug1: Connecting to 172.32.0.36 [172.32.0.36] port 61022. 15:57 < nng> debug1: Connection established. 15:57 < nng> debug1: identity file /home/pzapala/.ssh/id_rsa type 1 15:57 < nng> debug1: key_load_public: No such file or directory 15:57 < nng> debug1: identity file /home/pzapala/.ssh/id_rsa-cert type -1 15:57 < nng> debug1: key_load_public: No such file or directory 15:57 < nng> debug1: identity file /home/pzapala/.ssh/id_dsa type -1 15:57 < Psi-Jack> Please use pasteb 15:57 < nng> debug1: key_load_public: No such file or directory 15:57 < nng> debug1: identity file /home/pzapala/.ssh/id_dsa-cert type -1 15:57 < nng> debug1: key_load_public: No such file or directory 15:57 < junka> gg 15:57 < lopid> wd 15:58 < talx> you don't excess flood in freenode ? 15:58 < junka> you do 15:59 < Psi-Jack> Maybe one day he'll realize he flooded himself off IRC. :) 16:00 < blackflag_bfp> Anyone know where the AwesomWm people hang out? I went to #awesome on OFTC but that channel is dead quiet with no @ 16:00 < talx> Psi-Jack can I ask you a question ? 16:00 < Psi-Jack> blackflag_bfp: https://awesomewm.org/community/ 16:01 < BCMM> blackflag_bfp: check if there's also a freenode channel, maybe unofficial 16:01 < Psi-Jack> talx: You may. 16:01 < blackflag_bfp> Psi-Jack: thanks 16:01 < blackflag_bfp> BCMM: thanks to you as well 16:01 < BCMM> that happens for a few projects that are officially OFTC, due to people who can't be bothered connecting a 2nd network 16:01 < blackflag_bfp> I am having trouble making Awesome remember my nvidia-settings after reboot 16:02 < Psi-Jack> interrobangd: Daemonizing yourself? ;) 16:02 < talx> Psi-Jack: I've started a new linux course and the lecturer told us to send him what would like to learn 16:03 < nng> Psi-Jack: http://pasteb.com/5Ptf 16:03 < interrobangd> Psi-Jack, sorry :d .... login problem freenode 16:03 < Psi-Jack> nng: What about that ls? 16:03 < Psi-Jack> nng: And, yeah. Flooding == bad. Don't do it again. :) 16:03 < talx> Psi-Jack I've wrote things about grub sed and awk but do you have anything else that you might think would be important to know ? talking from expireince only 16:04 < nng> Psi-Jack: i know :) 16:05 < blackflag_bfp> I did check out that tmux. Pretty neat but a little redundant since i am already using awesome 16:05 < Psi-Jack> Depends. What are you trying to actually learn about, use Linux for, etc? 16:06 < talx> sysadmin 16:06 < maboc> blackflag_bfp: tmux rocks. especially because you can detach from the tmux. 16:06 < nng> Psi-Jack: kurahaupo ask :) sry 16:06 < blackflag_bfp> maboc: yeah I was lookng into that. Also it seems it works with ssh which I imagine awesome would not so well 16:06 < RevanOne> Anyone good with Bind ? 16:07 < Psi-Jack> nng: "sorry" not "sry" for future self corrections. 16:07 < maboc> @blackflag_bfp I use it on servers.....start a process...detach....come back later and reattach..... 16:07 < Psi-Jack> I asked for the output of a specific ls command that you have yet to provide. I never asked for what you provided. 16:07 < RevanOne> I am trying to setup a master - slave bind9. I am getting an error and the zones file are not created on the slave https://gist.github.com/dragosrosculete/26c1f0adab17caf5e03e857d3baccba1 16:08 < talx> Psi-Jack; it's System Admin 16:10 < silverdust> What's the linux equivalent of pbcopy ? i.e how can I copy to clipboard through command line 16:10 < LordRyan> xclip 16:10 < Psi-Jack> talx: Yep. So, you'd be good to learn things like Apache, nginx, haproxy, bind, postfix, dhcpd, osfp, bgp, iptables, iproute2, php-fpm, and.. Well, much much more. :) 16:11 < silverdust> LordRyan: Is that available in old linux boxes? 16:11 < talx> heh 16:11 < LordRyan> should be 16:11 < Psi-Jack> heh, pbcopy. Most macOS users don't even know about that, let alone people that don't even use macOS at all. 16:11 < LordRyan> i use it quite often 16:11 < LordRyan> unfortunately 16:12 < talx> thanks 16:13 < nng> Psi-Jack: http://pasteb.com/5RYG 16:13 < Psi-Jack> nng: Okay. Well. I don't have all day to wait for the requested information, so you'd best hurry up, or I'll give up waiting. :) 16:13 < Psi-Jack> Oh, finally. LOL 16:14 < Psi-Jack> That's the client system, trying to connect but failing? 16:14 < talx> Psi-Jack: what is osfp 16:14 < Psi-Jack> talx: Google, look it up. ;) 16:14 < talx> I don't find anything related on google 16:14 < Psi-Jack> Err, ospf. 16:14 < silverdust> I have a school server I ssh into to get stuff and restrictions are much on it. Turns out I'd have to install xclip from apt-get. So my options are either to find a way to do it with a POSIX command or find a way to install in my user directory withour requiring root permission 16:14 < nng> Psi-Jack: http://pasteb.com/5RYG 16:15 < Psi-Jack> nng: Yeah, you provided that paste already... 16:15 < JimBuntu> talx, When researching GNU, Linux/etc commands... it sometimes helps to add the word GNU or Linux or such in the google query 16:15 < nng> yes, so what would you like... ? 16:16 < Psi-Jack> An answer to the asked question relating to that. 16:16 < nng> maybe i don't read something... 16:16 < Psi-Jack> That's the client system, trying to connect but failing? 16:16 < Psi-Jack> Or the host you're trying to connect to? 16:17 < nng> yes that was from host where i will trying to connect 16:18 < Psi-Jack> Client? Or Destination? 16:18 < nng> destination 16:18 < Psi-Jack> Problem #1: authorized_keys should not be owned by root:root. 16:19 < Psi-Jack> chown user:group authorized_keys to the correct user and group. 16:19 < Psi-Jack> Second, I asked for the client side, NOT the destination side, but this provided useful information regardless. :) 16:20 < nng> client side : http://pasteb.com/5UBV 16:21 < Psi-Jack> There we go again, root group owned files. 16:22 < Psi-Jack> Fix those, as well. 16:22 < Psi-Jack> And avoid using root as if it were a regular user. It's not. :) 16:22 < nng> but if i cannect from regular user? 16:23 < nng> this same as in privilage? 16:23 < Psi-Jack> Your question is unclear. 16:23 < nng> whatever 16:26 < Hdphn> should I use debian testing 16:26 < Hdphn> or ubuntu 16:26 < Hdphn> I want my OS to stay away from my work 16:26 < Hdphn> just focus on my development 16:27 < sixardyH> ubuntu 16:27 < Hdphn> not debian testing? sixardyH 16:28 < Psi-Jack> Neither. 16:28 < Psi-Jack> :p 16:28 < sixardyH> i would say neither too, but between debian testing and ubuntu i'd pick ubuntu 16:28 < Psi-Jack> Debian testing is unstable, and Ubuntu... Is ubuntu. A disgrace to the entire Linux community many times over. 16:28 < nng> Psi-Jack: chown done, issue is still 16:29 < tadbitla> https://pastebin.com/raw/3rtnbmbq 16:29 < sixardyH> yeah but at least with ubuntu you shouldn't get any issues 16:29 < Hdphn> then what would you both recommend? Psi-Jack and sixardyH 16:29 < sixardyH> well uh 16:29 < Psi-Jack> Hdphn: I already told you what I recommended, yesterday. 16:29 < Hdphn> arch? 16:29 < Hdphn> I dont want that crap lol 16:29 < Psi-Jack> Hdphn: I already told you what I recommended, yesterday. 16:29 < sixardyH> as a gentoo user i'd recommend it but uh 16:29 < sixardyH> yeah 16:29 < Psi-Jack> It was not Arch, I never recommend Arch. 16:29 < sixardyH> i guess you could use something like opensuse 16:29 < sixardyH> or solus 16:29 < Hdphn> I want a decently tested OS. I dont wanna be tester myself. hence fedora and arch are out of the game 16:30 < Hdphn> sixardyH: isnt ubuntu more viable than opensuse or solus as it has big repo 16:30 < Hdphn> and widely famous 16:30 < Psi-Jack> Fedora is not you being a tester. 16:30 < Hdphn> its you being tester for 16:30 < sixardyH> well it depends on how familiar you are with linux 16:30 < Hdphn> ... 16:30 < Psi-Jack> Fedora is designed by developers for developers. 16:30 < Hdphn> RHEL 16:30 < Hdphn> :D 16:30 < Psi-Jack> Hdphn: Wrong. 16:31 < sixardyH> Fedora is a decent choice 16:31 < sixardyH> i just don't like rpm but that's just me 16:31 < sixardyH> i also prefer rolling releases 16:31 < Psi-Jack> No such thing as rolling "releases", because there is no "releases" 16:31 < Psi-Jack> :) 16:32 < JimBuntu> Amazon Linux X-D 16:32 < Hdphn> Psi-Jack: if you like fedora so much why did you switch back to arch ? :P 16:32 < Psi-Jack> Fedora does however, roll out updates for various common desktop software. 16:32 < ABCook> who has the best spin of Cinnamon ? 16:32 < Hdphn> fedora even implements SELinux in a hassle way 16:32 < Psi-Jack> ABCook: * 16:32 < Hdphn> unlike android 16:33 < Psi-Jack> Fedora's SElinux is actually pretty dang solid 16:33 < Psi-Jack> Very well configured, and comprehensive. :) 16:36 < NetTerminalGene> is selinux enabled by default in fedora? 16:36 < ayecee> i think so 16:36 < Hdphn> not anywhere near comparable to Android SELinux 16:36 < Hdphn> ;) 16:37 < sixardyH> install android x86 then 16:37 < sixardyH> /s 16:38 < NetTerminalGene> i was thinking using fedora but it has export law 16:38 < NetTerminalGene> it is subject to export law 16:39 < G3nka1> What I want to do is kill the java instances spawned from a particular jar (or a jar which contains a particular name) in bash. So I am planning to do jps -l and if the jar contains a string then pass that pid and do kill -9 on it. Is there a better way to do this? 16:41 < Psi-Jack> NetTerminalGene: Yeah, it uses SHA256 password hashes by default. 16:41 < Psi-Jack> And MozNSS FIPS compliant NSS and all that fun stuff. 16:42 < Psi-Jack> Hdphn: Android's SELinux is... Pretty crappy. 16:42 < Psi-Jack> Practically useless actually. 16:43 < ksk> G3nka1: never use kill -9 16:43 < G3nka1> ksk, then? 16:43 < ksk> and yes of course there are better ways, let the instance that start this thing stop it (Because it know pid, right?) 16:43 < ksk> like, for example, systemd? 16:43 < Psi-Jack> Not never. Just avoid it when you can. 16:44 < Psi-Jack> G3nka1: Again we don't support Windows here, too. 16:44 < P_B> kill -9 is my "Oh yeah! Well how do you like this, huh!?" option. 16:44 < Psi-Jack> You asked this last night then explained you were in git bash on Windows 10. 16:44 < NetTerminalGene> can i use aur with antergos? 16:44 < Psi-Jack> NetTerminalGene: Of course. 16:44 < NetTerminalGene> good 16:45 < tadbitla> P_B, Ctrl+Alt+Del alternative. 16:45 < ksk> NetTerminalGene: isnt everything subject to us export laws? should not interefere with your needs as long as you dont live in iran etc.. 16:45 < P_B> that always feels like defeat, tadbitla 16:45 < ksk> (everything as in every distro) 16:45 < P_B> What's the point if I just have to reboot? 16:45 < Hdphn> Psi-Jack: you kidding bro? 16:46 < Hdphn> Android 7.0 ++ has robust SELinux + other mitigations that no other linux distribution can offer as of yet. 16:46 < tadbitla> Well, the kernel doesn't panic if you grab the thread. It just resets to the login. 16:46 < Psi-Jack> Hdphn: No, I'm not. 16:46 < NetTerminalGene> ksk, it just makes me disturb 16:47 < Hdphn> archlinux doesnt even have MAC lol 16:48 < tadbitla> Psi-Jack, Everything is held accountable by a ledger who takes 2 names and names the highlight as a var. 16:48 < tadbitla> The 2 names get the credit, but the highlight just keeps working for the 2 names. 16:48 < tadbitla> I'll overthrow it like that. 16:48 < tadbitla> 2 names could be same name. So in this case, it's no coincidence. 16:49 < Psi-Jack> tadbitla: ... wut? 16:49 < tadbitla> That's why a lot of people are stuck. 16:49 < tadbitla> IRS and maybe NSA handle this. 16:50 < tadbitla> But I'm not politicing. 16:50 < Psi-Jack> More like rambling jibberish, 16:50 < ayecee> babbling 16:53 < xse> Hi, is it possible for dkms to redownload sources each time it needs to rebuild a module ? 16:53 < Psi-Jack> dkms doesn't download. 16:55 < xse> hm yeah that's what i though, i had issues with a module that wasn't building anymore since 4.15, it was fixed upstream tho it would have been nice for it to download that automaticaly :) 17:04 < blackflag_bfp> noob time with weechat I am having a hard time understanding how copy/paste works like if I wanted to cope a url into weechat to send in a msg 17:05 < sixardyH> so you want to copy an url from weechat? 17:05 < blackflag_bfp> no from system clipboard into weechat 17:05 < blackflag_bfp> and then I guess the reverse 17:05 < sixardyH> to paste in weechat, just do the same than in any terminal: Shift Ctrl V 17:05 < sixardyH> or shift right click -> paste 17:06 < sixardyH> some more niche terminals might not work like this though, idk 17:06 < blackflag_bfp> yeah that doesn't work for me so thats why im confused :( 17:06 < sixardyH> what terminal are you using? 17:06 < blackflag_bfp> sixardyH: Eterm 17:06 < sixardyH> never heard of it 17:06 < sixardyH> haha 17:07 < solidfox> on freebsd: man sudo has bad word wrapping (breaking words, and then line breaks right after for no reason 17:07 < blackflag_bfp> sixardyH: lol life with Linux right 17:07 < solidfox> but on ubuntu it looks great 17:07 < solidfox> why is that? 17:07 < sixardyH> blackflag_bfp: how do you usually paste in eterm? 17:07 < noodlepie> Linux kernel 4.16.1-gentoo stable here! 17:07 < blackflag_bfp> sixardyH: well I haven't figured that out yet lol wait one 17:07 < noodlepie> clear 17:07 < noodlepie> ifconfi 17:08 < noodlepie> oops 17:08 < sixardyH> nice 17:08 < blackflag_bfp> sixardyH: Shift + Insert 17:08 < JimBuntu> solidfox, does FreeBSD use the same GNU coreutils/etc as Linux? 17:09 < ayecee> normally no 17:09 < blackflag_bfp> Ahh ok that combination works in weechat as well. 17:10 < blackflag_bfp> Thanks! :) 17:10 < sixardyH> np :P that's a weird combination though 17:13 < solidfox> I think I'll stick with linux 17:13 < solidfox> freebsd is hard 17:13 < solidfox> gentoo is easier 17:13 < ayecee> :P 17:14 < blackflag_bfp> sixardyH: well I just built this monster 'WarMind" a few days ago so still playing around with different pkgs. Always open to recommended pkhs i.e. term,WM, editors. etc 17:14 < sixardyH> well uh do you prefer DE's or tiling wm's 17:15 < blackflag_bfp> well I am trying to learn the basics so my goal was to try and accomplish everything vi cli forst so I am running tiling "Awesome" currently 17:15 < Psi-Jack> sixardyH: Why not both? 17:19 < sixardyH> blackflag_bfp: if you like awesome and it fits for your workflow, then keep it. For terminals, I personally use xfce4-terminal (standalone), but most twm users prefer using urxvt, termite, or even st for the most minimalist users. If you want a CLI text editor, I would definitely recommend Vim : it's very hard to use at first, but the more you use it, the more efficient you will become. There are plenty of 17:19 < sixardyH> other useful CLI tools (tmux, etc) that you'll probably discover sooner or later, so don't hesitate to try them out and see if you like them 17:20 < blackflag_bfp> sixardyH: thanks for the info I just copied those down to checkout. I use Vim exclusively 17:22 < Psi-Jack> I use XFCE4 with i3wm instead of xfwm4, gnome-terminal, gedit, nm-applet, bluez, and all the trimmings. :) 17:22 * Psi-Jack takes his cake and eats it too. 17:23 < sixardyH> xfce is my favourite DE, but i prefer using standalone twm's 17:23 < sixardyH> i use herbstluftwm 17:24 < Psi-Jack> Pronounce that 5 times fast outloud without making a mistake./ 17:24 < sixardyH> it's german for autumn air 17:25 < dgurney> Psi-Jack, how well does xfce+i3wm setup work, and how hard is it to make? 17:25 < FreeFull> I use i3wm and a mix of various software 17:25 < sixardyH> also blackflag_bfp whenever you want to copy text from weechat but the UI gets in the way of your selection, use Alt+l to temporarily toggle text only mode, and copy your text 17:26 < Psi-Jack> dgurney: I use the whole xfce4-session management and everything, and not too difficult. There's an i3-workspaces-plugin for xfce4-panel which gives you i3's workspaces in the panel. Very handy. :) 17:26 < dgurney> neat 17:27 < Psi-Jack> dgurney: Beyond that it's just going into the xfce session manager, removing the xfwm4, adding i3 to the startup. 17:27 < Psi-Jack> You can reconfigure keyboard bindings within both XFCE and i3 as you deem appropriate. 17:27 < Psi-Jack> Not totally flawless integration, but it's pretty solid, and functional and beautiful. 17:28 < Psi-Jack> I use various xfce panel widgets for system stats. Much better than i3bar text-only blah. 17:28 < pankaj> I am learning assembly language. Can anybody suggest some latest debuggers for assembly language on linux destros? 17:31 < Psi-Jack> nasm, gdb, ddd. 17:31 < ayecee> gesundheit 17:31 < Psi-Jack> Heh 17:32 < sixardyH> i also use xfce4-panel actually 17:32 < sixardyH> was too lazy to configure polybar 17:33 < Psi-Jack> I simply... Don't want to setup polybar. 17:34 < pankaj> Psi-Jack: How to debug using nasm. 17:34 < pankaj> Psi-Jack: Allthough gdb is available it is not quite good. And ddd is not available. 17:34 < Psi-Jack> That, I couldn't help with. Good luck. :) 17:35 < Hdphn> Psi-Jack: please tell me 17:35 < Hdphn> why you switched from arch to fedora to arch again 17:35 < Hdphn> :D 17:35 < Psi-Jack> Those are the tools, though. :) 17:38 < nitenq> Hi how can I set manage-resolv-conf to false ? 17:38 < ayecee> do the opposite of how you set it to true 17:39 < nitenq> It's automatically set to true 17:39 < ayecee> says who 17:40 < nitenq> /etc/cloud/templates/resolv.conf.tmpl:# Your system has been configured with 'manage-resolv-conf' set to true. 17:41 < ayecee> so? 17:41 < nitenq> so how do I set it to false and where is it done 17:41 < ayecee> do the opposite of how you set it to true 17:42 < nitenq> wow thanks 17:42 < ayecee> np 17:43 < djph> ayecee: it's scary how "do the exact same thing except swap in 'false' for 'true'" needs to be stated 17:50 < Psi-Jack> Looks like cloud-init. 17:51 < ayecee> oh of course 17:51 < ayecee> what is cloud-init 17:51 < Pessimist> init but with clouds 17:51 < ayecee> makes sense 17:52 < Pessimist> ayecee, don't you love clouds? they are beautiful 17:54 < ayecee> often accompanied by rainbows too 17:55 < Psi-Jack> ayecee: It's automation for virtualization at the boot layer. Can pull in user-data to run at first boot, every boot, etc. 17:55 < Psi-Jack> Sets up the networking and more. 17:56 < ayecee> manages resolv.conf too, i guess 17:56 < Psi-Jack> AWS uses it, many VPS providers use it. Proxmox VE is implementing it, etc. 17:56 < Psi-Jack> It can. 17:56 < Psi-Jack> That's in line with setting up networking. :) 17:57 < Psi-Jack> it can provision networking from a 169.254.x IP, to a fully qualified IP+DNS etc, and even get user data from said 169.254.x IP 17:59 < JimBuntu> apipa (it's the same forward as backward) 18:00 < ayecee> upside down it's abiba 18:02 < chamarboii> hello guys I want 2 become a hacker helpp plzzz 18:02 < Psi-Jack> ayecee: No, upside down it's ɐdᴉdɐ 18:02 < ayecee> nice 18:02 < Psi-Jack> :) 18:02 < ayecee> i stand corrected 18:03 < rumpel> chamarboii, first you need a talking terminal 18:03 < Psi-Jack> chamarboii: ˙ǝuo ǝɯoɔǝq oʇ ʇɹɐʇs uɐɔ noʎ uǝɥʇ 'ɯɐɹƃoɹd oʇ ʍoɥ uɹɐǝ˥ 18:03 < OerHeks> chamarboii, if you can't do the time, ... 18:03 < sixardyH> Psi-Jack: you ruined my beautiful bitmap font with that text :( 18:04 < Psi-Jack> sixardyH: :) 18:04 < chamarboii> i know html 18:04 < ayecee> i know kung fu 18:04 < sixardyH> Times New Roman is my favourite programming language 18:05 < Psi-Jack> chamarboii: Do you have a linux question? This channel is about GNU/Linux, specifically, moderated, and not for aggresive uses, if that's what you think it is. 18:05 < ayecee> whatever happened to the original roman 18:05 < sixardyH> https://img.devrant.com/devrant/rant/r_730163_aaw94.jpg 18:06 < chamarboii> i am from korea .. my english bit week ... i try to learn linux for haacking 18:06 < chamarboii> i have a doubt 18:06 < ldiamond> Can you have a "mount pattern" in fstab? or with Systemd? I want to mount a bunch of files with a naming pattern into loopback devices (with same patter). 18:06 < ayecee> i would not think so 18:06 < Psi-Jack> chamarboii: What do you think "hacking" is? 18:07 < ayecee> it's not something people normally want to do with mount 18:07 < sixardyH> chamarboii: north or south? 18:07 < BCMM> ldiamond: it seems like that's an odd enough thing to do that there probably isn't built-in support for it, but it would be fairly straightforwards to just have a shell script that does it 18:07 < chamarboii> sixardyH south 18:07 < JimBuntu> Northies shouldn't be on here. 18:07 < ayecee> he's lying 18:07 < BCMM> ldiamond: (and have systemd run that shell script during boot) 18:07 < ayecee> i can tell 18:08 < Psi-Jack> chamarboii: What do you think "hacking" is? 18:08 < ksk> ldiamond: pam-mount can do things, not sure if you can trigger that withoug a login though. 18:08 < sixardyH> ldiamond: as BCMM said a shell script is your best option here 18:08 < ldiamond> ksk: login is fine 18:08 < ldiamond> yea a shell script sure, I hoped I could avoid the scripting part :p 18:08 < BCMM> chamarboii: out of curiosity, are you in korea or india? 18:09 < ldiamond> should be easy enough, probably only need 1 line of piping. ls | xarg | mount or something 18:09 < JimBuntu> whois herbert.freenode.net 18:10 < chamarboii> Psi-Jack: i am saw a movie about snowden .. i got inspired 18:10 < chamarboii> teach me 18:10 < chamarboii> guide me 18:10 < Psi-Jack> chamarboii: Uh huh. You are trolling. 18:10 < chamarboii> BCMM: I am from south korea, using vpn to hide from govt 18:10 < junka> troll me 18:10 < BluesKaj> named after frank herbert, a science fiction writer frpm the 50's and 60's I think, JimBuntu 18:10 < chamarboii> because hacking is illegal here 18:11 < Psi-Jack> !ops chamarboii Trolling 18:11 < junka> chamarboii; bb good luck with your life 18:11 < chamarboii> please noo 18:11 < JimBuntu> Thanks BluesKaj . If that's the case, that's an interesting naming scheme. 18:12 < chamarboii> i am getting error while trying to chmod /etc as root 18:12 < BluesKaj> JimBuntu, lots of freenode servers are so named, Isaac Asimov included :-) 18:13 < OerHeks> 42 18:14 < Psi-Jack> chamarboii: Too bad you had to troll. You might've actually gotten help. :) 18:15 < chamarboii> halp 18:15 < ABCook> a troll in chat? can't be so!! 18:15 < chamarboii> plzz 18:15 < ayecee> so rude 18:16 < balittad> acktually, it wanderful. 18:18 < chamarboii> existence is pain !ops Psi-Jack Harassment 18:18 < ayecee> i always forget how to spell that 18:18 < balittad> The Creatorian has risen. 18:21 < Psi-Jack> More jibberish? ;) 18:22 < chamarboii> ghftyhdkrtxdjrxd 18:22 < balittad> acktually, it's backward compatibility. :p 18:22 < chamarboii> oops bangs my head 18:23 < Psi-Jack> Welp. I hunger, time for lunch. :) 18:24 < chamarboii> Psi-Jack: u wasis 18:24 < chamarboii> Psi-Jack: u wasis basis 18:24 < chamarboii> Psi-Jack: u wasis basis crisis 18:24 < chamarboii> Psi-Jack: u wasis basis crisis lisis 18:24 < chamarboii> Psi-Jack: u wasis basis crisis lisis 18:24 < ayecee> please to stop 18:24 < chamarboii> Psi-Jack: u wasis basis crisis lisis chicken 18:25 < chamarboii> ok 18:25 < chamarboii> k 18:25 < chamarboii> i have an iq of 100 % 18:25 < chamarboii> i have an iq of 100 and fiddy 18:26 < ayecee> please to stop spam 18:26 < chamarboii> sucky sucky 10 bucky 18:27 < dchotas> i bursted out laughin at the sucky sucky thing 18:28 < Psi-Jack> Huh. Mobile irc client I'm using didn't even show the ban. But thank you sauvin. 18:28 < Dominian> get a better client 18:28 * Dominian ducks 18:28 < sauvin> Account was just a few minutes old. That one was a gimme. 18:29 * ayecee aims low 18:29 < Psi-Jack> mutter is an okay client so far. More for mobile use. Hehe 18:30 < Psi-Jack> sauvin: Yeah. I didn't think to check that. 18:31 < ayecee> so i'm looking at varnish for a reverse proxy to make separate servers available from a public address, and i'm starting to think this might be overkill. 18:31 < ayecee> certainly looks versatile though 18:33 < balittad> Swings with Chris Tuckings. 18:34 < balittad> But awdunno. 18:34 * ayecee peers at balittad 18:35 < mateothegreat> ayecee, haproxy will get you a lot of mileage for cheap 18:36 < noodlepie> Hi guys! 18:37 < Psi-Jack> Varnish can be nice but also can be overkill. 18:38 < ayecee> i think i might give it a shot just to get up to speed on it 18:38 < ayecee> also haproxy, thanks 18:38 < ksk> ayecee: Im using nginx for that. 18:38 < Psi-Jack> Haproxy actually can be interestingly utilized to reduce connections to a web application server. 18:39 < ayecee> i was just reading about how varnish does that too via request coalescing 18:39 < ksk> it does not have bling-bling overviews of backends in the open-source/free version, but who needs that for running a small-to-medium website 18:39 < ksk> nginx also has easy to use caching. 18:42 < ayecee> i'm using apache for that in a different circumstance 18:42 < Psi-Jack> For example. Haproxy -> nginx -> php-fpm 18:43 < Psi-Jack> Haproxy would be for tls termination, load balancing and connection pooling. 18:43 < Psi-Jack> Haproxy is better at it 18:44 < Psi-Jack> Haproxy can be configured to pool connections and drop connections upon completion so they don't stay hanging on when not needed. 18:49 < Psi-Jack> What problem are you trying to solve with Varnish, ayecee? 18:49 < ayecee> one public ip, multiple servers i want to reach 18:50 < Psi-Jack> Oh man. If that's all varnish is way overkill. Hehe 18:50 < ayecee> yup 18:51 < Psi-Jack> Definitely haproxy is more ideal for that. 18:51 < mateothegreat> SNI outta the box 18:52 < Psi-Jack> Varnish is not something that you can just drop in and start using. Often times it requires heavy changes to the backend to even benefit with what varnish can do. 18:52 < ayecee> varnish seems to have more capability to filter and transform requests too. 18:53 < Psi-Jack> Haproxy has a lot of filtering capability 18:54 < isutoshi> Hi everyone. For some reason mdadm on one of my servers has added an extra "ARRAY /dev/md/0 ..." line in my /etc/mdadm.conf. How can I do permanent changes to this file from within Busybox? 18:56 < Psi-Jack> Why busy box 18:57 < isutoshi> It's dropping to Busybox by itself, I didn't choose it. 18:58 < isutoshi> Forgot to mention that it's Debian Stretch, and a LUKS volume over a RAID1 array. 18:59 < hexnewbie> isutoshi: It's dropping you to initramfs because it can't mount /dev/md/0 or because it can't find root? The former would be really silly on its part 19:00 < isutoshi> It can't mount md0 saying that the passphrase is incorrect, but I think that's because it messed up my ARRAY lines. 19:01 < hexnewbie> isutoshi: I'd suggest you boot from Debian rescue mode of the netinst CD/flash (possibly live Debian if it can't mount the LUKS), then fix mdadm.conf and rebuild initramfs (update-initramfs -k all -u) 19:01 < hexnewbie> isutoshi: It sounds more like an actual issue with the passphrase 19:02 < isutoshi> I'll try booting off a USB drive, hexnewbie. Thanks! 19:04 < fr0tzed> sup 19:05 < backandforth> Hi, is it OK for my user is be part of the www-data group? 19:05 < fr0tzed> How do I point all logs to dev/null 19:09 < ephracis> Hi, anyone up for some network troubleshooting? Need some help. :) 19:10 < lf94> Hello ##linux. I'm trying to add a device id via echo "vvvv dddd" > /sys/bus/pci/drivers/radeon/new_id but I get "write failed file exists". Any ideas? 19:10 < jelly> lf94, what if you >> instead of > ? 19:11 < lf94> Same. 19:12 < ephracis> I switched router. Now Linux cannot ping the router or get an IP via DHCP, while Windows (dual boot) can. 19:12 < lf94> I have tried setpci too: setpci -vD -d 1002:9806 02.w=9802 19:12 < lf94> 1002:9806 is my graphics card (radeon hd6320), and I am trying change 9806 to 9802 19:14 < TJ-> lf94: " echo "vvvv dddd" | sudo tee /sys/bus/pci/drivers/radeon/new_id " 19:15 < lf94> TJ- I am super user already 19:15 < TJ-> lf94: okay, drop sudo then 19:15 < lf94> tee: /sys/bus/pci/drivers/radeon/new_id: File exists 19:15 < lf94> Also printed out what I echo'd in 19:15 < TJ-> lf94: that means the ID is already present 19:15 < TJ-> (the "file exists" message) 19:16 < lf94> Ok, so I guess I dont want to add a new id. 19:16 < lf94> I want to *change* what pci is reporting on my current device 19:16 < TJ-> the printing is what tee does (it T's the output to stdout and the named file) 19:16 < lf94> So setpci was the right direction here... 19:17 < lf94> # setpci -vD -d 1002:9806 02.w=9802 19:17 < lf94> 0000:00:01.0 @02 9802 19:17 < lf94> Is this a successful set? 19:32 < sirwilliam> Is there a way to bypass a public wifi that's blocking outgoing ssh on port 22? Without needing an external jump box. And the box I'm trying to ssh into only allows on port 22. 19:32 < BCMM> sirwilliam: no 19:33 < BCMM> sirwilliam: if you don't have anything on the outside that can listen on ports other than 22, and you can't send anything on port 22, you're out of luck 19:33 < sirwilliam> BCMM: thanks man, guess I will have to stick with using my cell's hotspot, lol. 19:34 < jelly> any kind of vpn, tunnel, or ssh-on-nondefault-port counts as "anything on the outside that can listen on ports other than 22" 19:35 < BCMM> jelly: i assumed that would come under "external jump box" 19:35 < BCMM> sirwilliam: if you can tether your phone on another interface, like usb, then you could set up a static route so that your ssh goes out over 22 and everything else goes over the free wifi 19:35 < BCMM> sirwilliam: (i'm assuming that the cellphone is charged per megabyte here) 19:35 < jelly> none of my boxes jump so dunno about that! 19:38 < sirwilliam> BCMM: Nice suggestion, no it's unlimited. 19:39 < sirwilliam> jelly: ha ha! I do have a paid VPN through ipvanish, how would one go about using that to tunnel ssh? 19:39 < ozymandias> turn on the vpn 19:39 < ozymandias> then ssh 19:39 < ozymandias> done 19:39 < ozymandias> :-D 19:40 < ozymandias> the traffic will be encrypted as it passes through the wifi, so they cannot inspect the ports it is going t 19:40 < ozymandias> to* 19:40 < ozymandias> and thus cannot block based on that 19:40 < sirwilliam> ozymandias: really that simple? Ha ha 19:40 < jelly> just like any other traffic passing through vpn, eg http or https 19:40 < sirwilliam> He he, I just had to ask the right people. 19:41 < ozymandias> could also ssh in off the phone, and add another listener port, and run it on more than one 19:42 < sirwilliam> Love it, thanks guys. 19:57 < linuxdood> Hi do we talk freebsd here or is that just for #freebsd? before i ask my questions. thanks 19:57 < Psi-Jack> linuxdood: ##linux is for Linux, ##freebsd is for FreeBSD. 19:57 < ayecee> linuxdood: freebsd isn't linux, so usually no. 19:57 < linuxdood> ok thanks! 19:57 < Psi-Jack> Here's your sign. :) 19:58 * ozymandias snorts 19:58 < codebam> so I'm using a terminal based irc client, how can I make my terminal display color unicode? 19:58 < codebam> I have the twitter emoji font installed 19:58 < Psi-Jack> Use a color unicode terminal. 19:58 < sauvin> You can talk "freebsd", but we'll answer "linux". If our Linux answers don't work for your FreeBSD situation, it's on *your* head. 19:58 < Psi-Jack> Configure it to use that font for xft. 19:59 < Psi-Jack> sauvin: Heh, he's already gone. :) 19:59 < o|0o^|> i used to like freebsd more than linux ban when i was a n00b 19:59 < codebam> Psi-Jack: how do I do that? I just put the emoji font after the current font in terms of priority? 19:59 < sauvin> I'm restating informal policy because I'm still sore about the dip that kept us going for hours about a grep question before fessing up he was on a Mac. 19:59 < ozymandias> you could probably uninstall the emjoi font 19:59 < Psi-Jack> That depends on many factors. Your terminal program, and your fontconfig. 20:00 < codebam> ozymandias: hmm? 20:00 < ozymandias> then you wont have to deal with it 20:00 < Psi-Jack> sauvin: Heh, ahh, G3nka1 was here doing it again earlier today, even. 20:00 < Psi-Jack> Err, thatw as Windows, different situation. ;) 20:01 < JimBuntu> and the one asking why the `man sudo` was wrongly formatted in FreeBSD but looked fine in ubuntu, may have been in #ubuntu 20:01 < codebam> Psi-Jack: um okay, I'm going to try to figure it out 20:01 < G3nka1> yup, had to figure out without #linux help although most of it was bash script. 20:01 < sauvin> There's a #bash if you need it. 20:01 < o|0o^|> you guys realize at least 66% of these people are trolls right 20:01 < G3nka1> sauvin, no fixed it, thanks though 20:01 < Psi-Jack> G3nka1: Good you figured it out,. Just remember, We support Linux, not bash script on Windows. :) 20:02 < sauvin> o|0o^|, never ascribe to malice what is more than adequately explained by ignorance. 20:02 < ozymandias> the point is not that people wont help you, its that you should share relevant details when asking for help 20:02 < codebam> 😄 20:02 < G3nka1> yes Psi-Jack 20:02 < ozymandias> many times people will still provide help, but it changes HOW they help, if they know something major is different than their assumptions 20:04 < sauvin> ozymandias is correct. 20:05 < oneplane> I want to decapsulate GRE traffic and forward the result on to another interface, but gre and gretap on linux don't see the incoming traffic at all, rcdcap wants ERSPAN and ipdecap doesn't work well on 64-bit systems :( 20:05 < ephracis> ozymandias: what info do you need? 20:06 < ozymandias> for what? 20:06 < ozymandias> in the example here, the assumption would be bash on linux, not bash on windows 20:06 < ephracis> I was asking for help, you said something about people helping when you give them enough info 20:06 < ephracis> ozymandias: oh, so you were not talking to me, then. :P 20:07 < sauvin> ephracis, the context in which that remark was made was before you walked in. :D 20:07 < ozymandias> :-D 20:07 < ephracis> sauvin: oh, I was here before, just had a disconnect :) 20:08 < ephracis> so I guess I am still out of luck with my network issue, then... :( 20:11 < JimBuntu> ephracis, if you manually configure your network in linux, does it then work... or not even then? 20:13 < ephracis> JimBuntu: I have tried with a static IP if that's what you mean. Still no luck 20:14 < ephracis> I am on Fedora 25, FYI 20:14 < ozymandias> isnt that EOL? 20:14 < JimBuntu> All you changed was the router? Is this wireless or wired? 20:14 < ephracis> wired 20:15 < ephracis> ozymandias: would love to upgrade, if I can just get the network running ;) 20:15 < ozymandias> fair enough 20:15 < ephracis> I can see ton of network from other hosts with tcpdump 20:16 < ephracis> and as I said, if I boot into Windows on the same machine with the same card/cable/router it works fine. 20:16 < JimBuntu> ephracis, for sanity, can you see if you can still connect to the old router with simple DHCP? 20:16 < ephracis> Sorry no, its left in the apartment I just sold. 20:16 < JimBuntu> My initial thought is that *maybe* you have the wired side for higher speeds only (new router) and the card doesn't go that high 20:17 < JimBuntu> Well, not with how it's set up in Fedora 20:17 < ephracis> But now I have an eth cable to my mac, trying to get that one working 20:17 < JimBuntu> Do you have a link light at all? 20:17 < ephracis> JimBuntu: yeah its connected alright 20:21 < Wulf> Hello. 20:22 < Wulf> How can I print a pdf file as "brochure"? E.g. pdf has 20 pages, I want 5 sheets of paper, folded in the middle. First sheet would have pages 1,20 on the front, 2,19 on the back, etc. 20:26 < LissajousPattern> hi 20:26 < hexnewbie> Wulf: Use pdfbook, pdfnup, etc. from texlive-utils-extra (that's the Debian package, don't know the name on other distros) 20:26 < hexnewbie> Wulf: It's command-line-only, unfortunately, and a bit difficult to wrap your head around 20:29 < Wulf> hexnewbie: pdfbook2 does the job, thank you! 20:39 < ephracis> Nope, I am lost at this. Guess I am just gonna reinstall everything, hope that helps. :) 20:40 < JimBuntu> ephracis, maybe a bootable USB first? Of course, if you are re-installing, you can move up and out of 25 I guess too 20:40 < balittad> Reboot, then re-install, then re-boot then login. Then re-boot, then re-install. 20:40 < TJ-> ephracis: is it using an Intel PCI ethernet interface? 20:41 < TJ-> ephracis: because what you describe sounds like a common problem with the chipset not waking up the transmit side of the PHY, due to the Windows driver setting power-saving modes and Wake-on-LAN on the interface. 20:45 < ephracis> TJ-: Yeah, on the motherboard 20:45 < ephracis> JimBuntu: Yeah, that was my plan :D 20:46 < ephracis> TJ-: So you're saying Windows is again giving me headaches and anger? :P How can I check/fix? 20:47 < Psi-Jack> Remove Windows, use only linux. Pain solved. 20:48 < JimBuntu> ^^ +1 20:48 < TJ-> ephracis: check the device type: "lspci -nn -d :0200" and pastebin the result 20:48 < zer0G> hello, need urgent help 20:49 < solidfox> zer0G, ok 20:49 < solidfox> zer0G, start downloading sysRescueCD 20:49 < infinisil> zer0G: can't help without a problem 20:50 < solidfox> infinisil, +1 20:50 < zer0G> when you read rc like "~/.vimrc" and "~/.bashrc", is it pronounced Vim-rick (like limerick) or like R-C (as in Init R C) 20:50 < TJ-> ephracis: we've dealt with this several time in #ubuntu over recent months, always the same: a certain Intel ethernet chipset and Windows dual boot - the window's driver sets these power-save options which the Linux driver cannot undo when there's a warm reboot. A clean cold boot from a system that has been powered off (at the wall - not just left on 5V standby) will usually clear it 20:50 < solidfox> wtf 20:50 < solidfox> thats not urgent lol 20:50 < solidfox> I always say R C 20:50 < busybox42> lol 20:50 < zer0G> sorry, sense of urgency is not my strong suit 20:50 < busybox42> I alawasy say "bash r c" 20:50 < JimBuntu> runcom files 20:50 < zer0G> nor is pronunciation 20:51 < ephracis> TJ-: cool, gonna give it a go 20:51 < infinisil> /filter add zero * nick_zer0G * 20:51 < infinisil> Oh sry ignore that 20:51 < ayecee> ignore what 20:51 < JimBuntu> tilde whack dot bash r c (IMHO) 20:51 < solidfox> /ignore solidfox 20:52 < solidfox> whoops 20:52 < zer0G> lol, thanks for the help ~ 20:52 < solidfox> didnt mean to 20:52 < solidfox> JimBuntu, I always say "noodle" instead of tilde 20:52 < solidfox> and slash 20:52 < Ben64> i say "squiggly" 20:52 < solidfox> noodle slash dot 20:52 < JimBuntu> solidfox, however you wanna salt your meal is fine with me :-) 20:52 < ayecee> i usually reserve whack for backslash 20:53 < ayecee> forward slash is just slash 20:53 < solidfox> % grapes, ~ noodle, : salt 20:53 < solidfox> sorry salt is :: 20:53 < Ben64> a lot of people call / backslash 20:54 < solidfox> Ben64, my teacher in highschool did that. drove me nuts 20:54 < ayecee> a lot of people believe the earth is flat 20:54 < Ben64> idk about a lot 20:54 < solidfox> ayecee, cuz it is 20:54 < ayecee> that's what they want you to think 20:54 < solidfox> lol jk 20:54 < ayecee> teach the controversy 20:54 < Ben64> hollow sphere earth 20:54 < ozymandias> solidfox, go build a steam powered rocket already ;-) 20:54 < ozymandias> or buy a plane ticket and get higher :-D 20:55 < ephracis> TJ-: wow! It fixed it! That's great, thanks! 20:55 < solidfox> Ben64, the earth is more like a blanket 20:55 < ephracis> Now onto my next issue: 100% CPU from gnome-shell ever since I changed GPU to GeForce 1060. Haha! :) 20:55 < zer0G> dodecahedron earth makes more sense... then nobody is wrong -- kinda flat, kinda round 20:56 < solidfox> ephracis, what the heck 20:56 < TJ-> ephracis: :) you can fix it permanently by changing the Windows driver options for that device, in Device Manager. The tab that has the power saving options on 20:56 < TJ-> ephracis: configured to use the wrong drivers, or a mix of drivers? 20:56 < Psi-Jack> Heh. 20:57 < Psi-Jack> Fix permanently by completely oblitering Windows from your existance. 20:57 < Psi-Jack> It's the only way to be sure. :) 20:57 < ozymandias> JimBuntu, the dot in the letter 'i' is called a 'tittle' 20:58 < TJ-> Psi-Jack: it's actually a bug in the Linux driver, not correctly initialising the device 20:58 < ayecee> the more you know ~* 20:59 < Psi-Jack> TJ-: But it only surfaces when running Windows, no? 20:59 < Psi-Jack> Or dual-booting? ;) 21:00 < phogg> People who call / "backslash" seem to me to be confused and think that there is no difference between / and \ but it's more technical sounding to call it by the longer word. 21:00 < JimBuntu> a / leans forward, \ leans back. Seems easy to remember. 21:01 < phogg> if you pronounce / as anything other than "slash" you're also missing out on the joke in Slashdot's name. 21:01 < phogg> JimBuntu: unless you're in an RTL language! 21:01 < JimBuntu> nah, got the joke 21:01 < ephracis> TJ-: thanks, I will check it the next time I'm in Windows. You're golden! 21:01 < TJ-> Psi-Jack: it surfaces because the windows driver puts the chipset into a power-saving mode correctly but on boot the Linux driver only sets the Rx power saving, not the Tx, so the device can see packets but nothing gets transmitted out 21:01 < Psi-Jack> phogg: What joke in Slashdot's name? 21:01 < JimBuntu> phogg, that's a good point... but then again "backslash" isn't written in a right to left language ;-) 21:01 < ephracis> Not sure what the GPU issue is. Gonna start by updating the system now when I have network again 21:01 < TJ-> Link is up, everything looks great, ethtool can't change it 21:01 < phogg> Psi-Jack: it's funny when you read the URL aloud 21:02 < Psi-Jack> Ahhhhh. Yes. 21:02 < JimBuntu> I always thought dot-slash would have been a better one 21:02 < phogg> JimBuntu: maybe the name was taken 21:02 < JimBuntu> dot.com 21:03 < ozymandias> dotcom.com 21:03 < phogg> "The owner of dotslash.org is offering it for sale for an asking price of 1749 USD!" 21:03 < phogg> nobody wants that domain as badly as all that 21:03 < ayecee> it's a steal at twice the price 21:03 < Psi-Jack> Nope, dotslash.org registered in 1998, after the fact of slashdot.org in 1997. 21:03 < aeyxa> what is a reason for umount hanging but then works with -l ? 21:03 < aeyxa> like terminal would freeze 21:03 < ozymandias> sometrhing is using it, aeyxa 21:04 < ozymandias> -l is lazy --- as in, unmount when things are done 21:04 < phogg> quick, somebody ask taco what he was thinking! 21:04 < aeyxa> but it instantly unmounted with that 21:04 < ozymandias> and stop new things from accessing it until then 21:04 < ozymandias> no it didnt 21:04 < Psi-Jack> Has slashdot always been SourceForge? 21:04 < aeyxa> when I reran mount it was gone? 21:04 < ozymandias> it just looks that way 21:04 < BCMM> didn't that used to be a slashdot parody site, back in the day? 21:04 < BCMM> dotslash, i mean 21:04 < aeyxa> alright 21:04 < phogg> aeyxa: it instantly returns to the prompt; actual unmounting will happen some time later (or never) 21:05 < aeyxa> I ran mount though and it's not listed is that also a possible lie? 21:05 < ozymandias> yes 21:05 < ozymandias> it is misleading 21:05 < ozymandias> potentially 21:05 < phogg> aeyxa: it was removed from the list of mounted things, but it's not necessarily actually unmounted. 21:05 < ozymandias> since the process that was using it may still be using it 21:05 < aeyxa> how should I make sure it's closed? it's clearly doing something I don't want 21:05 < ozymandias> kill the process using it 21:05 < ozymandias> or stop it cleanly, if you can 21:06 < JimBuntu> try running a manual sync first 21:06 < aeyxa> it's an nfs mount, I'm not sure what's using it 21:06 < ozymandias> check with lsof 21:06 < phogg> aeyxa: there are two things going on: a list of what is mounted and actually busy devices. If the device is busy you can't actually dereference it, but you can remove it from the list. Doing so is not wise since you then can no longer track when it's really no longer in use. Sometimes you need to, but be aware it's not safe. 21:07 < phogg> typically you use umount -l when you know it's safe anyway 21:11 < Psi-Jack> Got a CPU core always pinned at 100% utilization? Have you checked your system for the new trend of hostile cryptocurrenty mining? heh 21:13 < ayecee> what a world we live in 21:14 < Psi-Jack> Right? Heh. $85k worth of manero (sp?) coin per day... 21:14 < ayecee> monero, like money 21:14 < ayecee> and dinero 21:14 < Psi-Jack> Your Android phone running hot? Check it for a bit coin miner, else it might explode, literally. 21:14 < phogg> like dinner 21:15 < ayecee> en espagnol 21:21 < alexey-nemovff> cripto-minería, ¿en qué mundo vivímos? xD 21:22 < aeyxa> I think maybe the problem with the nfs mount is that the folder is inside a git repo so when I git pulled changes it might have somehow gotten the nfs stuck? 21:23 < ayecee> sure 21:23 < Psi-Jack> Folder? File Cabinet? Manilla? Or colored? 21:23 * Psi-Jack grins. 21:23 < o|0o^|> i store all my gits in a trapper keeper 21:23 < Psi-Jack> heh 21:23 * ayecee hwacks Psi-Jack with a manila folder 21:23 < notmike> My device has been rek'd since the 4.15 kernel. I have an AMD GPU, and I suspect the DC/DAL patch is at least contributing to the problem. What happens, 1) boots normally to xfce login screen, 2) after login, upon suspend (usually) I lose all display and network connectivity. 21:24 < notmike> I want to test if amdgpu.dc=0 kernel option fixes it, but I don't know where to inject kernel options. In grub? 21:24 < o|0o^|> why would amdgpu shut down your network? 21:25 < notmike> I don't thin it's related, but it does occur so I mentioned it. 21:25 < ayecee> notmike: /etc/default/grub is where i do it 21:25 < notmike> Literally 4.14 works flawlessly. 4.15+ rek'd my shit 21:26 < o|0o^|> if you don't update your user space can't break *taps the ole noggin* 21:27 < notmike> Well, I reverted to 4.14 but I need to get this issue fixed. I'm two updates behind. 21:27 < ayecee> :o 21:28 < notmike> The AMD patch is just the most obvious change that was made. 21:28 < notmike> And I do have an AMD GPU 21:28 < o|0o^|> blacklist the driver and see if it keeps happening 21:29 < notmike> Won't that make my gpu not work? 21:30 < o|0o^|> will it? i don't know shouldn't there be some fallback mode? 21:30 < o|0o^|> like VESA or VGA or someth'n 21:30 < notmike> I'm not knowledgeable about that. Where do I blacklist? 21:31 < o|0o^|> some conf file i forget 21:31 < o|0o^|> modprobe.somethingorother 21:31 < Psi-Jack> Heh. 21:31 < Psi-Jack> Blind leading blind. :) 21:34 < deepfreez> Is anyone here with buildrpm exprience? I want to create a php rpm with my config etc.. Any ideas? 21:34 < notmike> It's /etc/modprobe.d/ I think 21:34 < Psi-Jack> Do not ask to ask. 21:34 < ayecee> ask if you can ask to ask first 21:34 < ananke> deepfreez: use an existing source rpm. replace the config. build it 21:36 < Psi-Jack> ^ Very good advice to start from. 21:36 < o|0o^|> i think theres a kernel cmdline that modprobe will look for too, but don't remember that either 21:36 < Psi-Jack> deepfreez: What are you really trying to do though, actually? There's likely a much better way. :) 21:37 < deepfreez> To compile the php in my way 21:37 < Psi-Jack> deepfreez: Why? What's the purpose of that? 21:38 < deepfreez> to install fast php? 21:38 < qoxncyha> is it meTAcity or MEtaCIty? 21:38 < Psi-Jack> deepfreez: That makes no sense. What are you really trying to do? What is your real end goal? 21:38 < deepfreez> ... 21:39 < anonnumberanon> man this is weird can i get help? 21:39 < anonnumberanon> i try to use Tail 21:39 < anonnumberanon> so i make a file 21:39 < o|0o^|> and then??? 21:39 < anonnumberanon> do this next: $ tail -f 21:39 < Psi-Jack> anonnumberanon: Hitting enter as pauses in thought and/or punctuation makes things annoying. 21:39 < phogg> qoxncyha: It's not pronounced 'meta city', the A is as in "bat" and is stressed. Metacity" 21:39 < anonnumberanon> and on the tail terminal when i change something in the file and hit save, i don't see it in the tail terminal 21:40 < o|0o^|> anonnumberanon: in a text editor of some sort? 21:40 < o|0o^|> it's probably replacing the file, not adding any data to it 21:40 < anonnumberanon> yes just in Mousepad 21:40 < qoxncyha> phogg: can you point me to a reference for that? 21:40 < phogg> qoxncyha: No. I forget where I learned this. 21:41 < phogg> qoxncyha: Might have picked it up talking to people who *hopefully* had it right. 21:41 < anonnumberanon> o|0o^|, how do you ONLY add data to a file? 21:41 < qoxncyha> thanks 21:42 < o|0o^|> echo "blahh" >> thefile 21:42 < Psi-Jack> deepfreez: Well? What is your real actual goal here? I'd like to help you, but re-inventing the wheel unnecessarily isn't helping. 21:43 < deepfreez> I want to create a rpm with my of configure / etc... to install on multiple servers? 21:43 < deepfreez> rpm is more faster like ./configure... etc 21:43 < o|0o^|> if you strace the text editor you'll see exactly what it's doing, either renaming or unlinking depending on how oblivious the authors were 21:43 < Psi-Jack> deepfreez: Sounds like what you need is configuration management. RPM is not intended for configuration management. 21:44 < deepfreez> you do read what you type? 21:44 < Psi-Jack> Yes, I read every single word. 21:44 < deepfreez> if you don't know to answer my question, don't do it... 21:44 < deepfreez> because you are far far away... 21:45 < Psi-Jack> I know the answer, it's you that doesn't know the proper ways to do what they actually want properly. :) 21:45 < deepfreez> sorry but I think you don't know how to do it. 21:45 < Psi-Jack> Been at this over 24 years. :) 21:45 < deepfreez> ok 21:45 < Psi-Jack> This CentOS, RHEL, or SUSE? 21:45 < deepfreez> CentOS 21:45 < Psi-Jack> Okay. 6 or 7? 21:46 < deepfreez> both 21:46 < deepfreez> But right now I try on 7 21:46 < axd-v> I got a new computer and I simply moved my ssd from my old one to the new one (x230 thinkpad) and debian booted without a problem as if this ssd has always been native. However, are there some drivers and/or optimizations that are missing since I didn't install this os physically on this computer? 21:46 < Psi-Jack> Why are you compiling a custom build of PHP specifically? 21:46 < deepfreez> Because I need custom compile 21:46 < al2o3-cr> deepfreez: don't e so fucking cheeky, people here are trying to help. 21:47 < Psi-Jack> deepfreez: The Remi Repos for PHP are very good, and made by people whom are employed at Red Hat Software. Do they not fit the needs better than custom compiling your own? 21:47 < Psi-Jack> al2o3-cr: Kindly mind the language, :) 21:47 < al2o3-cr> sure. 21:47 < deepfreez> ok 21:48 < deepfreez> how I use a remi srpm to be compiled in my own way? 21:48 < Psi-Jack> ... 21:48 < Psi-Jack> What is specifically different about YOUR OWN WAY? 21:48 < Psi-Jack> Seriously. 21:48 < deepfreez> I want to be installed in /opt/ 21:49 < deepfreez> and other stuff 21:49 < deepfreez> wtf 21:49 < Psi-Jack> Remi uses /opt 21:49 < deepfreez> stop 21:49 < deepfreez> you don't know how to do it 21:49 < Psi-Jack> I've just proven many times over that I do. 21:49 < deepfreez> I know about remi etc.. but the result is not ok 21:49 < deepfreez> ok 21:49 * lupine installs deepfreez into /opt 21:49 < Psi-Jack> Your bull-headedness is what's making you think unclearly and lash out aggressively. 21:49 < deepfreez> close the subject. 21:50 < lupine> anyway, ./configure --prefix=/opt 21:50 < lupine> you're welcome and so on 21:50 < deepfreez> How I create my own rpm from php srpm? this is the question 21:50 < o|0o^|> --prefix=/opt/php , you animal! 21:50 < deepfreez> lupine, you are beside the subject 21:50 < fendur> :> 21:50 < phogg> ugh, /opt )-: 21:50 < Psi-Jack> Some people, refusing to do things good. *shrugs* 21:51 < phogg> Psi-Jack: are we confusing good and right again? Or no? 21:51 < deepfreez> Yap :) none use rpmbuild 21:51 < Psi-Jack> Both, actually. :) 21:51 < deepfreez> to compile his own rpm 21:51 < o|0o^|> oh you're making a package 21:51 < deepfreez> yes I want to create my own package :) 21:52 < o|0o^|> ./configure --prefix=/install/to/here/on/target/system && make && DESTDIR=/put/package/files/here make install 21:52 < Psi-Jack> LOL 21:52 < phogg> that's not quite all there is to it 21:53 < deepfreez> LOL 21:53 < o|0o^|> tar -cJf thefiles.tar.xz package.tar.xz 21:53 < Psi-Jack> I'm still wondering, what's "not good" about Remi PHP builds. ;) 21:53 < phogg> fortunately for me I have blanked the bits of my memory covering this due to immediate and dedicated consumption of alcohol 21:53 < o|0o^|> fucked it up 21:53 < o|0o^|> I FAILED YOU ALL! 21:53 < Psi-Jack> o|0o^|: Ahem, language. :) 21:53 < phogg> otherwise I might feel obliged to help 21:54 < phogg> o|0o^|: you need neither - or J 21:54 < fendur> o|0o^|: 2 days in the dungeon. 21:54 < Psi-Jack> Hung by toe clamps. 21:55 < fendur> I was gonna say he can take his laptop. sheesh! 21:55 < o|0o^|> oh yeah don't forget to tar -tf thefiles.tar.xz > thefiles.pkg so you don't forget what you just installed 21:55 < phogg> sentence commuted to one day in the dungeon if that time is spent composing an apology poem in iambic pentameter 21:55 < fendur> or she, I dunno., 21:56 < Psi-Jack> deepfreez: You really don't want to explain why you need to reinvent the wheel specifically? 21:56 < deepfreez> Psi-Jack, I just want to learn how to...because all know how but in practice... the failed.. 21:56 < phogg> Psi-Jack: remi PHP packages have problematic dependencies. I had to build my own due to a need to use openssl instead of libnss. 21:56 < Psi-Jack> deepfreez: You might start, then, with a simpler need. ;) 21:57 < Psi-Jack> phogg: I use Remi PHP repos, and have for years, they have no issues. 21:57 < deepfreez> ok 21:57 < Psi-Jack> OpenSSL however vs NSS, well, CentOS uses nss at its core of the OS itself. Sometimes can be annoying, but I've never seen that issue come up specifically for PHP. 21:57 < phogg> Psi-Jack: I am happy for you. I assure you that if there were an option I would have used the binaries, but there was not. 21:57 < deepfreez> If you want to create a rpm form remi ( to custom with / enable ) how you do it? 21:58 < Psi-Jack> deepfreez: Download the src.rpms'. 21:58 < Psi-Jack> Start there. 21:58 < phogg> Psi-Jack: I had a PHP application which made assumptions about being run with openssl and would not accept NSS no matter how nicely I asked. 21:58 < Psi-Jack> phogg: Hmmm. 21:59 < phogg> Psi-Jack: rebuilding PHP was by far easier than trying to get the app fixed. If it were just that one thing I might have fixed it anyway, but there were a few other issues as well. 21:59 < phogg> that's just the one I remember 22:00 < Psi-Jack> phogg: Yeah. I had a similar issue, but not with PHP, instead openLDAP. Needed to use pbkdf2 with openldap, and the plugin for that depended on openssl. Obviously openldap from EL6 was built against NSS, and wouldn't work. :) 22:00 < phogg> php sadly has a ton of build time options and not everyone agrees on what should be built. Enable internationalization via libintl? Some apps require it, but often people don't care. 22:00 < phogg> Psi-Jack: because of that one PHP issue I had to get custom copies of libcurl, curl and Apache as well. It was not a fun time. 22:00 < Psi-Jack> I understand that PHP 7 is working to fix all those kinds of issues. 22:01 < phogg> Psi-Jack: Theoretically. They suffer from the main devs being idiots, and that's still hurting them no matter the version. 22:01 < Psi-Jack> heh 22:01 < phogg> Psi-Jack: I ran in to a new one today: Subclasses of SPL don't respect __debugInfo 22:02 < phogg> so if you extend e.g. \ArrayObject you can't control it's debug output. 22:02 < Psi-Jack> phogg: Was funny though, when I first joined the project where openssl was being recompiled for pbkdf2... Originally it was a chef recipe that compiled it during provisioning, on the host. :) 22:02 < phogg> eww 22:02 < Psi-Jack> I came in, fixed it up, made the openssl-project specific openldap replacement in rpm. :) 22:02 < phogg> why so late? 22:03 < phogg> we still build everything non-standard and bundle it into a single package 22:03 < Psi-Jack> Now, I need to re-leard a bit of how to build .deb packages, despite my utter putrid hatred of it. ;) 22:04 < phogg> .deb is not so bad, it's just the documentation is not all it could be 22:04 < Psi-Jack> Thankfully, I expect not to need to build custom ones of those as much. But, I do have some things. 22:04 < Psi-Jack> .deb is horrible. 22:04 < Psi-Jack> It was horrible 22 years ago, it's worse today than it was back then. :) 22:04 < phogg> I will never agree. Building RPMs is enough to drive me mad. Utter junk from front to back, the whole process. 22:05 < Psi-Jack> The only part that I don't like about rpm is organizational . 22:05 < phogg> Meaning what? 22:06 < zer0G> meaning the rpm blows up your directory 22:06 < Psi-Jack> ~/rpmbuild/{SPECS,SOURCES,etc} is monolithic. 22:06 < lupine> deb > rpm 22:06 < lupine> in literally every way 22:06 < lupine> (including the only important one, which is standards of maintainership) 22:06 * Psi-Jack pulls the level releasing the trapdoor under lupine's feet. 22:06 < phogg> Psi-Jack: yes, that is specifically what I hate about it 22:06 < Psi-Jack> lever& 22:07 < solidfox> do you guys use tmux 22:07 < Psi-Jack> phogg: That's the only thing about it that could be better. :) 22:07 < phogg> Psi-Jack: with .deb you have SO MUCH more flexibility 22:07 < Psi-Jack> solidfox: Vote somewhere else. Do you have an actual question? ;) 22:07 < phogg> Psi-Jack: At least you're not insane, then (-; 22:07 < lupine> all technical considerations pale before the simple fact that debian does a much better job of maintaining its repositories than RH-alikes 22:07 < solidfox> Psi-Jack, no I don't. 22:07 < solidfox> Psi-Jack, lol 22:07 < Psi-Jack> phogg: heh. At the same time, I hate that .deb uses a debian directory INSIDE the source tree. 22:08 < phogg> lupine: That stops being true as soon as you start looking at debs from vendors or PPAs. Same problems there. 22:08 < lupine> yeah, I wouldn't look at those 22:08 < lupine> quality is highly variable in third-party repos on both sides 22:08 < busybox42> Has debian figured out how do to differentials in packages updates yet? 22:08 < phogg> Psi-Jack: Why? The original source stays in a separate tarball in any case. It's better than having to somehow associate a spec file. 22:08 < Psi-Jack> busybox42: Nope. 22:08 < busybox42> :) 22:08 < phogg> differentials meaning what? 22:09 < phogg> apt has been doing binary diffs for over a decade 22:09 < busybox42> meaning you pull down just the files that changed. 22:09 < Psi-Jack> phogg: Delta packages, like drpm. 22:09 < phogg> Psi-Jack: delta packages no, but delta bytes on the wire yes 22:09 < Psi-Jack> deltarpm's are very nice for reduced bandwidth requirements. 22:10 < phogg> I remember when that feature landed. Saves tons of bandwidth. 22:10 < busybox42> I don't like most things about debian based distros but it's probably because I've been working with RedHat so long I'm just bias. 22:10 < Psi-Jack> Even opensuse uses delta-rpms now. 22:10 < busybox42> And Ubuntu is kind of a mess. 22:10 < lupine> yeah, I wouldn't touch ubuntu :p 22:10 < phogg> busybox42: Probably. Just about everything in Debian is well thought out. You may not always agree, and I sure don't, but it's pretty carefully planned. I appreciate that. By contrast even RHEL seems haphazard. 22:10 < lupine> the technical considerations of the two formats are not important. what matters is the skill with which the collections of packages are maintained. and debian, specifically debian, does really well in that metric 22:10 < hanetzer> install.gentoo.meme ;) 22:11 < Psi-Jack> lupine: Debian still has /etc/init.d scripts for things that should be using systemd service units. 22:11 < solidfox> Psi-Jack, I can ask for opinions here 22:11 < solidfox> Psi-Jack, don't you like tmux? 22:12 < Psi-Jack> I don't like people taking polls, which are useless for all. 22:12 < solidfox> Psi-Jack, I'm just curious! 22:12 < phogg> Psi-Jack: https://debian-administration.org/article/439/Avoiding_slow_package_updates_with_package_diffs 22:12 < phogg> Psi-Jack: I don't know why you think apt doesn't do this. That's 2006. 22:13 < lupine> Psi-Jack: the migration isn't complete, for sure 22:13 < lupine> but it's smooooooooooooth 22:13 < Psi-Jack> phogg: Hmmm, Is that enabled by default? 22:13 < phogg> Psi-Jack: yes 22:13 < lupine> upgrades go really well in debian. ubuntu hasn't managed to replicate that 22:13 < lupine> it throws in do-release-upgrade to try to paper over the cracks 22:14 < Psi-Jack> I will give Debian that. It upgrades, and dist-upgrades insanely well. 22:14 < jimm> Psi-Jack, most of those have been rewritten to call the systemd equiv, if systemd is installed 22:14 < phogg> I disagree. Everything should upgrade at least as well as Debian. If your system isn't up to it just use apt and dpkg. It's a solved technical problem we should stop trying to re-invent poorly. 22:15 < Psi-Jack> jimm: But, not all. 22:15 < lupine> apt and dpkg are not why debian upgrades well 22:15 < lupine> ubuntu proves that 22:15 < lupine> it's the careful maintainership that makes it possible 22:15 < phogg> lupine: Debian policy is why 22:15 < lupine> ubuntu doesn't replicate that, and neither does redhat 22:15 < lupine> aye 22:15 < ozymandias> hey now 22:15 < ozymandias> RHEL is phenominal at upgrades, too 22:15 < Psi-Jack> ozymandias: No, it's really not. 22:16 < lupine> good one 22:16 < ozymandias> fedora was not for a long time, but RHEL has been rock solid 22:16 < phogg> on the internet no one can tell when you're joking 22:16 < phogg> I assume it's a joke anyway 22:16 < Psi-Jack> Fedora is the first one to FINALLY be able to dist upgrade to newer versions, and it's in part due to dnf. 22:16 < jimm> phogg, some folks doing deriv dists don't know how to handle the entirety of debian 22:16 < phogg> Psi-Jack: Really? They stopped recommending a clean install? 22:16 < Psi-Jack> phogg: Really. 22:16 < phogg> Psi-Jack: Nice to see progress there. 22:17 < Psi-Jack> dnf based dist-upgrading is VERY impressive, and solid. 22:17 < Psi-Jack> it downloads everything then does a reboot to install everything, and then reboots into the upgraded system. 22:17 < phogg> ah, still requires a forced reboot... no thanks 22:17 < Psi-Jack> CentOS/EL, still recommends a full new install. 22:17 < Psi-Jack> phogg: You should reboot... 22:17 < phogg> maybe they'll catch up for RHEL8 22:18 < phogg> Psi-Jack: No. I'll decide when I should reboot. 22:18 < Psi-Jack> Yes, you do decide. When you want to do a major upgrade, you actively choose to reboot to do so. ;) 22:18 < phogg> I can get 90% of the usefulness of the new version without rebooting. Only kernel, drivers and some lower level bits can't change online. 22:19 < phogg> Psi-Jack: There is no reason to postpone the entire update until a reboot. 22:19 < Psi-Jack> Not to mention everything using glibc, which is everything. :) 22:19 < phogg> Psi-Jack: No. Debian will upgrade glibc online just fine. 22:19 < Psi-Jack> Sure, but you're not USING the new glibc until restarting everything that had the old one loaded. 22:19 < phogg> Psi-Jack: it warns you about which daemons need to restart and prompts to restart them 22:19 < jimm> I left redhat (rhel and fedora didn't exist yet) because it couldn;t dist-upgrade, stayed with debian because it consistantly could 22:20 < Psi-Jack> phogg: Daemons aren't the only thing on your system, too. :) 22:20 < phogg> Again, very few changes require a reboot. If your apps require new kernel features or new udev features they may not work, but that's rare. 22:21 < phogg> Psi-Jack: and I am perfectly happy to upgrade what can upgrade and then wait until a good time to reboot. I still get most of the update's usefulness immediately: I can spawn new processes at new versions. 22:22 < Psi-Jack> At great cost for stability, definitely. 22:22 < phogg> Psi-Jack: At a theoretical cost to stability. I know where the risks are and I reboot if there's one I can't risk. 22:23 < dgarstang> How do I remove a sysctl setting from memory? 22:24 < hanetzer> dgarstang: if you set it as sysctl set foo.bar.baz 1, then do sysctl set foo.bar.baz 0 or whatever the syntax is 22:24 < dgarstang> hanetzer: this is a string. When I use /sbin/sysctl -w kernel.hostname="" I get error: Malformed setting "kernel.hostname=" 22:25 < phinxy> Does a NFS shared directory need special care if its a root filesystem that will be booted from? 22:27 < dgarstang> hanetzer: Any idea? 22:27 < o|0o^|> dgarstang: might work with " ", but then it tells me cannot stat /proc/sys/hostname no such file or directory 22:28 < dgarstang> hanetzer: I also tried echo "" > /proc/sys/kernel/hostname but then the hostname becomes '' 22:28 < dgarstang> o|0o^|: Well this sucks 22:28 < dgarstang> seems a reboot is required? But ... this is Linux, not Windows 22:30 < ayecee> )`: 22:30 < phogg> dgarstang: what are you trying to do? 22:30 < dgarstang> phogg: unset kernel.hostname from memory 22:32 < phogg> dgarstang: setting the file in sys does that. The hostname becomes an empty string. 22:32 < dgarstang> phogg: Setting it in sysctl? 22:32 < dgarstang> /sbin/sysctl -w kernel.hostname='' == 'error: Malformed setting "kernel.hostname="' 22:33 < o|0o^|> no idea why sysctl is looking in /proc/sys instead of /proc/sys/kernel 22:33 < Psi-Jack> Ugh. Can't believe I'm copying 3 months worth of verified spam back from archives into the active trap account. :) 22:33 < phogg> dgarstang: echo "" > /proc/sys/kernel/hostname does what you are asking for. 22:33 < cluelessperson> I'm getting 400Mbps with an AP I know to support 1300. How do I go about troubleshooting this? 22:34 < dgarstang> phogg: echo "" > /proc/sys/kernel/hostname && python -c 'import socket ; print socket.gethostname()' ... gets me an empty hostname 22:34 < ayecee> reorient the receiver with respect to the transmitter 22:34 < ayecee> like the FCC says 22:34 < phogg> dgarstang: yes, that's what you asked for 22:34 < dgarstang> phogg: Before this kernel.hostnam was set, running python -c 'import socket ; print socket.gethostname()' would return the correct FQDN 22:35 < dgarstang> phogg: ANY setting of that overrides the correct FQDN 22:35 < phogg> dgarstang: what does hostname -f report? 22:35 < dgarstang> phogg: The correct value. However, php and python and splunk all return whats in kernel.hostname 22:36 < dgarstang> phogg: Correction... 22:36 < dgarstang> phogg: if it's set, that is what is returned. 22:36 < dgarstang> It looks like a reboot of ALL of my servers is required 22:36 < dgarstang> Seems you can't UNSET a sysctl variable 22:37 < phogg> dgarstang: Try > /proc/sys/kernel/hostname without the echo 22:37 < phogg> might differentiate based on the newline 22:37 < dgarstang> phogg: nah that didn't do it 22:38 < dgarstang> python -c 'import socket ; print socket.gethostname()' still returns what was in there before, it what was set in sysctl 22:38 < pankaj> Hello, I serched about a torrent client named frostwire but is it old or still in wide usage among linux users? 22:39 < dgarstang> phogg: Actually running " > /proc/sys/kernel/hostname" doesn't overwrite what's already in the file 22:39 < phogg> dgarstang: interesting 22:39 < hipp> pankaj> i use rtorrent 22:39 < phogg> pankaj: I only use rtorrent 22:39 < dgarstang> > /proc/sys/kernel/hostname && cat /proc/sys/kernel/hostname ... = 'w1' 22:39 < hipp> ^^ 22:39 < phogg> pankaj: if you need a GUI I understand transmission is OK too 22:40 < anonnumberanon> o|0o^|, actually tail kind of works now. still testing 22:40 < ayecee> oh, that's someone's nick 22:41 < phogg> ayecee: what is 22:41 < ayecee> was having a hard time parsing the emoticon 22:41 < pankaj> phogg: I just want to download an older version of ubuntu (needed for use of a legacy application) mainly 9.10 Karmic Koala, but torrent for it are not availble. 22:41 < ayecee> phogg: the one that looks like an emoticon 22:42 < kfrench> ayecee: I think it's a foot with an infected toe in a sandal. 22:43 < ayecee> i can see that 22:43 < kfrench> You can't unsee that 22:43 < jkemppainen> UH 22:43 < jkemppainen> well, this is an interesting conversation to walk into ... 22:43 * jkemppainen shifty eyes 22:43 < ayecee> welcome to the internet. you must be new here. 22:45 < phogg> Only the first trauma is free. After that you must traumatize someone else in return. 22:45 < ayecee> PTSD it forward 22:50 < phogg> hmm. sysctl is perhaps the dumbest program I have ever seen 22:50 < ayecee> NO U 22:52 < phogg> ayecee: It literally reads and writes files in /proc/sys in the simplest way you can imagine 22:53 < ayecee> pretty much, yeah. also sanitizes the environment while doing it. 22:54 < ayecee> i think 22:54 < phogg> doesn't look that way 22:54 < phogg> I'm surprised it even bothers to swap / for . and back again. 22:54 < ayecee> ¯\_(ツ)_/¯ 22:54 < ayecee> could be a convention that comes from another platform where sysctl is used. 22:55 < phogg> ayecee: it's okay you can say "Solaris" 22:56 < ayecee> it would feel weird, since i was thinking freebsd. 22:57 < Dagmar> TRIGGERED 23:05 * Psi-Jack squeezes the trigger. 23:07 < Psi-Jack> Ohhh... I miss the days when Dominian would randomly show up during a scenario like that, bury the body in follow-through, and be on his way again. :) 23:08 < ayecee> every day i'm dumping the bodies 23:08 < Psi-Jack> All the bodies that are hitting the floor? ;) 23:09 < Psi-Jack> Now, it was funny. One day at my former job, we suddenly had a delivery for 1,000 body bags. 23:09 < diogenese> they must be piling up by now 23:09 < ayecee> it's raining men, hallelujah it's raining men 23:10 < diogenese> take cover 23:10 < Psi-Jack> Heh, one of our offices happened to be right next to the city coroner's office. :) 23:10 < ayecee> handy 23:10 < Psi-Jack> Was even funnier when the company employees referred to themselves as "Mobsters" 23:13 < ibttis> do you guys know of a good tutorial on how to secure a newly installed linux system, harden it against potential attacks? 23:13 < Dagmar> If it were so easy it could be put into a tutorial, it woudn't be a multi-million dollar industry 23:14 < Psi-Jack> ibttis: Security is hard. Not one tutorial's going to teach you something so complex. 23:14 < Dagmar> Always apply your patches. 23:14 < Psi-Jack> Patches. heh. 23:14 < ibttis> well or a good tutorial on iptables, tcpdump, nmap etc. 23:14 < Psi-Jack> nmap? That's random. 23:15 < Psi-Jack> ibttis: What distro? 23:15 < Dagmar> As a complete non-sequitir, I'm having a *very* good day today. I got a call from a credit fraudster whose hang-up button appears to not work 23:15 < Psi-Jack> hah 23:15 < Dagmar> He's now listening to some gay bear porn Google was nice enough to find for me in an incognito search 23:15 < ibttis> ubuntu Psi-Jack 23:15 < Psi-Jack> Oh... 23:16 < Psi-Jack> Dagmar: But, doesn't that mean... YOU are also listening to said gat bear porn? 23:16 < Psi-Jack> gey* 23:16 * Psi-Jack destroys keyboard, replaces with new. 23:16 < Dagmar> nope. I plugged the laptop into the wall and the phone into the laptop in the garage about 20 minutes ago 23:16 < Dagmar> I just checked on them 23:16 < Dagmar> Such language he is using now 23:17 < Psi-Jack> Dagmar: Heh, seriously? They cannot hang up? 23:17 < Dagmar> If he could hang up he would have surely done so by now. 23:18 < Psi-Jack> heh, must be a big glitch in the software they're using. 23:18 < Dagmar> I practically did a binary search of "things that are upsetting to hear" on the guy for the first ten minutes 23:18 < Dagmar> He should have never let on that he was homophobic. Tsk tsk 23:18 < Psi-Jack> Well, the longer you keep them on the line, the less they get to try to call others. :) 23:19 < Dagmar> That's nice, but I'm really enjoying the automated suffering of the guilty. 23:21 < Dagmar> In a few more minutes i"m going to go back in there and see if he's still around and suggest that the reason he hasn't hung up is that he's really quite enjoying himself 23:23 < dysfigured> if i'm using lvm can i shrink my /home partition while i'm currently using it (or dismounting somehow) without rebooting into a live image? or is it better to do 'from orbit'? 23:23 < ayecee> it depends which filesystem you used 23:23 < Psi-Jack> Shrinking, you don't want to do live. 23:23 < dysfigured> ext4 23:23 < ayecee> i don't think ext4 can be shrunk online 23:23 < Psi-Jack> It cannot. 23:24 < dysfigured> damn. well, what file systems would let you? 23:24 < ayecee> that's probably why i think that! 23:24 < Dagmar> dysfigured: Filesystems that aren't really concerned with data loss 23:24 < ayecee> dunno 23:24 < Psi-Jack> There was one..... I just don't remember which one. 23:24 < Psi-Jack> JFS? 23:24 < Psi-Jack> But,... JFS.. Meh. 23:24 < dysfigured> weelll i kinda like the idea of data integrity.. so i guess nvm on that 23:25 < Psi-Jack> "nevermind" not "nvm" for future self correction. 23:25 < dysfigured> ookay 23:25 < Psi-Jack> It matters. :) 23:25 < Dagmar> Now that 4Gb thumbdrives are stupid-cheap, I just keep distros on them for booting up in situations like that 23:28 < Psi-Jack> I have a multi-boot USB that is scripted very well to detect what ISO's are on the drive and creates a structured menu of them all. :) 23:28 < dysfigured> mm that sounds nice 23:28 < Dagmar> Very 23:28 < mallu> I'm trying to create a systemd service which copy file from the server before shutting down. How can I ensure network service is available while the script run? 23:30 < Psi-Jack> It's quite nice. The power of grub2 is actually quite impressive. 23:30 < Psi-Jack> if conditions, loops, basically think sh scripting. 23:30 < Psi-Jack> But omit the normal CLI tools you might have. :) 23:32 < Psi-Jack> And now.... All swords sharpened, time to head home. 23:33 < rockdarko> Can I ask an apache2 related question in here? Will I got shot? 23:34 < lupine> seems unlikely 23:34 < ayecee> wouldn't rule it out though 23:34 < lupine> but I'll admit to firing up my puzzlement generator already 23:34 < ayecee> besides, we normally stick to lynching 23:34 < rockdarko> weehaa 23:35 < rockdarko> I used letsencrypt tog enerate my conf files for my virtual hosts but as much as they seem to be properly configured to redirect any http request to https it doesn't. Anyone good with that? 23:36 < ayecee> good with what 23:36 < ayecee> troubleshooting unseen config files? 23:36 * ayecee gets the rope 23:36 < lupine> perhaps it's generating for the wrong version of apache? 23:36 < lupine> 2.2 vs. 2.4 23:36 < rockdarko> ayecee: haha 23:37 < rockdarko> lupine: maybe, I'll see if that's possible. But for whats it wrong my ssl conf file is there: https://darko.vision/pub/httpdvconf01 and my non-ssl one is there: https://darko.vision/pub/httpdvconf01-nonssl. 23:37 < rockdarko> for what its worth* 23:38 < ayecee> this rewrite rule doesn't match the hostname 23:38 < lupine> god, mod_rewrite 23:38 < lupine> indeed it does not 23:38 < rockdarko> ayecee: because its using the root host name rather than the sub domain? 23:39 < ayecee> yeah 23:39 < rockdarko> darko.vision vs owncloud.darko.vision ? 23:39 < ayecee> yes 23:39 < rockdarko> ok thanks, I was wondering that, but the file was generated by letsencrypt so I thought it was odd, it's not like I wrote it, servers me right I guess. 23:39 < rockdarko> Thanks so much. 23:40 < no26> Hello. I search what handler take care of my hardware interrupt on Linux. I know I can see via /proc/interrupt the handlers binded with the irq, but I need to see the effective handler who are called. It's possible ? 23:42 < ayecee> those all look like words, but i don't know what you're asking 23:47 < jgarr> Someone know a way I can log how much traffic is used per process? I found plenty of information per interface and nethogs will give me an interactive view of the data but I want historical data so I can troubleshoot a problem that happens when no one is logged onto a system. 23:48 < xz> jgarr, I guess you would end up monitoring socket activity that way 23:48 < jgarr> Enabling connection tracking (/proc/net/nf_conntrack) is the best I've found but it shows bytes per connection, not per process. I could cross reference the data from /proc but was curious if a tool exists already 23:49 < kurahaupo> jgarr: multiple processes can be writing to or reading from a socket, so accounting per process can be quite tricky 23:49 < Hanumaan> I have 3TB HDD in which there were 2 btrfs partitions (Top 2 with names in the given link: Backup1, Backup) and a FAT32 partition after Analyse I have got this result with Testdisk. Can Testdisk recover partions and data or is there any other better tool? Also I need FAT32 parition : https://paste.ubuntu.com/p/qCj3WgpJRy/ 23:50 < xz> kurahaupo, I thought socket was exclusive to the process that created it 23:50 < mallu> Can someone please tell me how I can run a command during shutdown but before network service stop? 23:51 < kurahaupo> xz: no, it can be shared with children, just like any other filedescriptor 23:51 < xz> mallu, do you use systemd? 23:51 < mallu> xz: yes 23:51 < xz> kurahaupo, ok, but then you can monitor just parent process 23:51 < xz> mallu, you might help on #systemd 23:51 < jgarr> kurahaupo: do you know of another place that information might be stored? /proc/$pid/net/dev is system info, not per process 23:52 < mallu> join /systemd 23:52 < xz> /join #systemd 23:52 < mallu> oops... thanks xz :) 23:52 < kurahaupo> xz: it's also possible to send a filedescriptor between two cooperative but unrelated processes, but not very common. 23:53 < xz> jgarr, doesn't wireshark offer any sort of smart filter that would allow you for per process monitoring? 23:53 < jgarr> nethogs gets the data but I haven't (yet) dug into their code to see how they do it 23:53 < xz> jgarr, I'm not sure, however, if wireshark knows about sockets 23:54 < kurahaupo> xz: the networking stack does record the UID of the sending process, for inspection in the OUTPUT chain 23:55 < kurahaupo> (of iptables) 23:55 < xz> kurahaupo, does wireshark catch that? or they just sniff bare data on the interface? 23:55 < kurahaupo> Wireshark and tcpdump use the same kernel interface 23:55 * jgarr is looking through tcpdump man page --- Log closed Wed Apr 11 00:00:40 2018