| === markthomas is now known as markthomas|away | ||
| === Lcawte is now known as Lcawte|Away | ||
| jrwren | start -v mongodb says start: Job failed to start | 02:16 |
|---|---|---|
| jrwren | there are no logs anywhere that I can find. Where to look? | 02:16 |
| sarnold | try /var/log/upstart/mongo* ? | 02:17 |
| jrwren | nothing in /var/log/upstart/ nothing in /var/log/syslog | 02:17 |
| jrwren | sarnold: I wish :( | 02:17 |
| Patrickdk | just run mongo manually and see what happens | 02:17 |
| jrwren | Patrickdk: works great :( | 02:18 |
| sarnold | jrwren: you could try running fatrace to see what files are accessed | 02:18 |
| jrwren | even invoking with start-stop-daemon the way the upstart job does works great. | 02:18 |
| Patrickdk | you are running it as the same user as the init script uses? | 02:18 |
| jrwren | fatrace is a good idea. | 02:18 |
| jrwren | start-stop-daemon has a user option, so yes? | 02:18 |
| jrwren | even adding debug to init/mongodb.conf gives no log | 02:21 |
| sarnold | anything in dmesg? | 02:21 |
| sarnold | I understand if the numa memory allocation policy isn't set correctly for mongo it'll just fall over with an oom | 02:22 |
| jrwren | nothing obvious. this is in an lxc and so dmesg is showing me host stuff. Still, nothing with a recent timestamp | 02:23 |
| jrwren | and mongo runs fine if I invoke it manually. upstart seems to be the problem? | 02:24 |
| shellox | hi | 02:25 |
| sarnold | oh man that pokes a neuron.. | 02:26 |
| jrwren | sarnold: there is an issue with running mongo in a host if its already running in lxc, but this is not that :( | 02:26 |
| shellox | I'm using a ubuntu server 14.04, nginx and php-fpm and was wondering where the best place is to install a php application? One developer will need access to it using sftp, so i was thinking /srv/www/my-app and give him access to this directory | 02:27 |
| sarnold | jrwren: ah! that's what it was. thanks.... | 02:27 |
| shellox | is that a sound way to do it :P? | 02:27 |
| sarnold | shellox: yeah, it's as good a place as any. debian guidelines might suggest stuffing it somewhere under /var/www but I never really liked that location for locally installed web applications. | 02:28 |
| sarnold | jrwren: but dang. :) | 02:28 |
| jrwren | is there a way to put upstart into super debug mode? | 02:30 |
| jrwren | or /sbin/start into debug verbose++ mode? | 02:31 |
| shellox | sarnold: Is it good practice to add a user to the www-data group, so he has write access to directory? | 02:31 |
| sarnold | shellox: I'd rather have the directory and files owned by the developer, and make sure the www-data user or group can read the files. I think webservers should only have write access to their log files and databse sockets... | 02:33 |
| Patrickdk | webservers shouldn't have to write to them ever | 02:33 |
| Patrickdk | your cgi/fcgi might have to, and should be limited | 02:33 |
| Patrickdk | it's nice to do something like git or so | 02:34 |
| Patrickdk | and just have it continuously get updates pushed to it | 02:34 |
| Patrickdk | no need to worry about dev owning files | 02:35 |
| jrwren | strace of the /sbin/start is a bit interesting | 02:36 |
| jrwren | http://paste.ubuntu.com/11216946/ | 02:36 |
| Patrickdk | not really, to be expected | 02:37 |
| Patrickdk | as strace doesn't follow forks | 02:37 |
| jrwren | yeah? the EAGAIN on reads to the /com/ubuntu/upstart socket is to be expected? | 02:38 |
| Patrickdk | can't tell, it's all ... | 02:39 |
| Patrickdk | but I don't see that | 02:39 |
| jrwren | i'll paste again with -f and -s 1024 | 02:39 |
| Patrickdk | it sent a message, it worked | 02:39 |
| shellox | sarnold: I see, that sounds plausbile. Thanks | 02:40 |
| jrwren | http://paste.ubuntu.com/11216993/ | 02:40 |
| Patrickdk | looks like it's working fine | 02:41 |
| sarnold | I didn't suggest strace since I expected it to be useless; iirc start just contacts the running pid 1 and asks -it- to start a new job, and upstart uses ptrace extensively when starting new jobs, and .. throw containers inthe mix and I really doubt strace will demystify anything. | 02:42 |
| jrwren | ok. maybe trace pid 1? | 02:43 |
| Patrickdk | no | 02:43 |
| sarnold | I haven't got a clue if that's allowed, and I wouldn't try it on a system I needed to stay alive :) | 02:43 |
| Patrickdk | you need to trace the mongodb pid | 02:43 |
| jrwren | mongo is fine. | 02:43 |
| jrwren | there is no mongo pid | 02:43 |
| Patrickdk | sure there is | 02:43 |
| jrwren | if I start mongo manually, it runs | 02:43 |
| jrwren | if I start mongo by using start-stop-daemon exactly as is in the upstart script, it runs. | 02:44 |
| jrwren | upstart is failing. | 02:44 |
| sarnold | maybe throw the strace -f on that start-stop-daemon | 02:44 |
| jrwren | but that actually works. | 02:44 |
| jrwren | oh, on it in the upstart job? | 02:44 |
| sarnold | again, since upstart uses ptrace I'm worried it might not be helpful... | 02:44 |
| jrwren | oh! | 02:48 |
| jrwren | this is a user lxc | 02:48 |
| jrwren | and mongodb upstart jobs has | 02:48 |
| jrwren | limit nofile 64000 64000 | 02:48 |
| jrwren | which failed | 02:48 |
| jrwren | strace -p 1 showed me. | 02:49 |
| jrwren | that was it. | 02:49 |
| sarnold | woo! | 02:49 |
| jrwren | HUGE thanks for helping. | 02:49 |
| jrwren | i don't know if that is a bug or not. | 02:49 |
| sarnold | it seems worth filing a bug report | 02:49 |
| jrwren | ok. | 02:49 |
| sarnold | but I don't know if it ought to be filed against lxc, mongo, or upstart :) | 02:49 |
| jrwren | ha! that was about to be my next question | 02:50 |
| sarnold | heck file it against all three ... | 02:50 |
| jrwren | i'll have to learn how to LP to file 1 bug against all 3. | 02:50 |
| sarnold | file it against one package, then hit the "also affects distribution/package" button, and add in another one in the 'source package name' field | 02:51 |
| jrwren | thanks. | 02:51 |
| sarnold | time to bail, have fun jrwren :) | 02:53 |
| jrwren | thanks again. good night. | 02:53 |
| === gema_ is now known as gema | ||
| === markthomas|away is now known as markthomas | ||
| hxm | morning | 06:05 |
| hxm | i had some problems with mysql, so i unninstalled every mysql packages, i am trying to install mysql-common and it says /etc/mysql/my.cnf.fallback does not exists | 06:07 |
| hxm | indeed, i deleted that directory | 06:07 |
| === markthomas|away is now known as markthomas | ||
| rbasak | hxm: purge mysql-common first. Otherwise it'll keep your "modifications". | 07:42 |
| rbasak | hxm: in fact, make sure to purge *all* MySQL related packages if you're going to go to the extreme of removing stuff manually first. | 07:43 |
| === Lcawte|Away is now known as Lcawte | ||
| shellox | which permission should the key file for my SSL cert have? | 08:29 |
| shellox | i found this guide | 08:29 |
| shellox | https://help.ubuntu.com/lts/serverguide/certificates-and-security.html | 08:29 |
| shellox | and copied it to /etc/ssl/private | 08:29 |
| === ashleyd is now known as ashd | ||
| Sander^work2 | In what circumstances is it required to have nfs shares inside a subdir like this: /nfs/host1/share1 and nfs/host2/share2/ instead of having everything in /nfs/share1 and /nfs/share2 | 10:13 |
| === ashleyd is now known as ashd | ||
| === ashleyd is now known as ashd | ||
| === ashleyd is now known as ashd | ||
| === ashleyd is now known as ashd | ||
| === ashleyd is now known as ashd | ||
| === ashleyd is now known as ashd | ||
| === ashleyd is now known as ashd | ||
| hxm | I have this problem http://paste.ubuntu.com/11223995/ which is a headache for me | 11:50 |
| hxm | in the /var/log/mysql/error.log the relevant line is ERROR: 1062 Duplicate entry 'innodb' for key 'PRIMARY' | 11:52 |
| hxm | i thought i already deleted database | 11:53 |
| === ashleyd is now known as ashd | ||
| === ashleyd is now known as ashd | ||
| === ashleyd is now known as ashd | ||
| === ashleyd is now known as ashd | ||
| === ashleyd is now known as ashd | ||
| === ashleyd is now known as ashd | ||
| === ashleyd is now known as ashd | ||
| === ashleyd is now known as ashd | ||
| === ashleyd is now known as ashd | ||
| rbasak | matsubara_: around? Would you mind driving SRU verification for bug 1443735 for me please? It'll be a while yet - an existing SRU needs to clear before I can upload it and it'll probably sit in the SRU queue for a while after that anyway, but I thought I'd ask you in advance. | 14:07 |
| rbasak | https://bugs.launchpad.net/ubuntu/+source/grub2/+bug/1443735 | 14:07 |
| * rbasak wonders where the bot is. | 14:07 | |
| * matsubara_ looks | 14:10 | |
| matsubara_ | rbasak, ok. I can do that | 14:13 |
| rbasak | matsubara_: thanks! | 14:15 |
| Voyage | how to write automated scripts for linux with max possibilities and control? | 14:26 |
| jrwren | Voyage: very carefully. | 14:27 |
| Voyage | jrwren, in what way | 14:27 |
| jrwren | Voyage: that is how I write automated scripts. I do it very carefully. | 14:27 |
| Voyage | you know scripts that would do stuff that I do in commandline. rsync ssh, copy files, change configs, connect to another server, do stuff, return back. what ever....... | 14:28 |
| Voyage | how to do it? | 14:29 |
| Voyage | how do you write those | 14:29 |
| Voyage | jrwren, bash or what? | 14:30 |
| Voyage | I heard doing this with pythong was a smarter way | 14:30 |
| jrwren | Voyage: bash is fine. its nice to have easy access to gnu coreutils things. | 14:31 |
| jrwren | Voyage: I recommend you start with bash and use something else only if you need it. | 14:31 |
| Voyage | cant python do what bash can? | 14:31 |
| jrwren | Voyage: I love python too. I'd not call it "a smarter way" | 14:31 |
| Voyage | why | 14:31 |
| jrwren | Voyage: it depends. | 14:32 |
| jrwren | Voyage: you'll find in tech, a lot of people have strong opinions and like to tell others what they should be doing without any idea of individuals actual needs. | 14:32 |
| jrwren | Voyage: because many things can be done simply in bash which aren't quite as simple in python, or... require some external python dependency. | 14:33 |
| Voyage | hm | 14:33 |
| === ashleyd is now known as ashd | ||
| Voyage | can anyone give an example of things that cannot be done by bash and we need python to do so? | 14:35 |
| jrwren | Voyage: there are no absolute answers to these qeustions. | 14:37 |
| jrwren | Voyage: python can spawn threads, bash doesn't have primitives for that | 14:37 |
| jrwren | Voyage: BUT, python has its GIL, so threads aren't as useful. | 14:37 |
| jrwren | Voyage: I highly recommend focusing on solving problems isntead of worrying about differences between python and bash. | 14:38 |
| Voyage | hm | 14:40 |
| hackeron | hey, anyone has an issue with Ubuntu 15.04 not rebooting without a connector monitor? - Any ideas what could be causing it? -- 14.10 rebooted without issues. | 14:45 |
| hackeron | (I'm using ubuntu-server, so no X and I would like to run headless) | 14:46 |
| === ashleyd is now known as ashd | ||
| rbasak | hackeron: https://bugs.launchpad.net/ubuntu/+source/grub2/+bug/1443735 maybe? | 14:53 |
| rbasak | You should be aware of that and implement the workaround on every headless box anyway. We should have an update which fixes the default soon. | 14:54 |
| hackeron | rbasak: no, not that one - I already have this set on every headless server | 14:54 |
| hackeron | rbasak: this is related to 15.04 and I am guessing maybe systemd - plugging in a monitor makes the server instantly boot | 14:54 |
| rbasak | hackeron: that's interesting | 14:55 |
| rbasak | hackeron: if you figure it out, please let me know. Maybe try a serial USB adaptor to diagnose if that works? | 14:56 |
| hackeron | rbasak: I commented on that original bug back in 2011 | 14:56 |
| hackeron | rbasak: the server is remote somewhere, I am speaking to them on skype and when I reboot the server, it doesn't boot - as soon as a monitor is plugged in, it boots up - nothing in the logs other than:May 19 15:45:20 TimeBox rsyslogd: [origin software="rsyslogd" swVersion="7.4.4" x-pid="1055" x-info="http://www.rsyslog.com"] exiting on signal 15. | 14:57 |
| hackeron | May 19 15:53:14 TimeBox rsyslogd: [origin software="rsyslogd" swVersion="7.4.4" x-pid="1653" x-info="http://www.rsyslog.com"] start | 14:57 |
| hackeron | the rsyslogd start happens the instant a monitor is plugged in - and this did not happen in 14.10 | 14:58 |
| === ashleyd is now known as ashd | ||
| rbasak | hackeron: nothing in dmesg? | 15:26 |
| hackeron | rbasak: nope, the kernel starts booting the instant the monitor is plugged in - it appears to get stuck in grub | 15:26 |
| rbasak | Ah | 15:26 |
| rbasak | The only logging you'd get from grub is to the console really (practically that means serial) | 15:27 |
| hackeron | trying to boot with nomodeset - but not sure how that will help, if it never starts booting without a monitor :/ | 15:28 |
| hackeron | rbasak: don't have a serial console though - what do they look like these days anyway? - I only have USB ports, heh | 15:29 |
| rbasak | hackeron: yeah I understand. Nowadays the only form I'm aware of are USB<->serial adaptors. I have no idea if grub can use one of those though. | 15:31 |
| hackeron | rbasak: even so, what do I connect to the serial end of the adapter? | 15:31 |
| rbasak | hackeron: a smartphone video of the screen is another possibility, but also impossible here because your screen won't be plugged in | 15:31 |
| rbasak | hackeron: another usb serial adaptor :) | 15:31 |
| hackeron | rbasak: the second I plug a monitor in, it boots :/ | 15:31 |
| rbasak | Yeah | 15:31 |
| rbasak | VGA? | 15:32 |
| hackeron | rbasak: yep | 15:32 |
| rbasak | The only other thing I can think of is to cut the DDC pins with a switch for debugging | 15:32 |
| rattking | I alot of OOB managment systems do SOL, some ever provide access via ssh.. very nice compared to some java kvm | 15:33 |
| rbasak | It seems a bit of a stretch though I think DDC is the only way grub would know when the monitor gets plugged in | 15:33 |
| === ashleyd is now known as ashd | ||
| hackeron | rbasak: hmmm, if I set all the GRUB timeouts to 0 (effectively disabling the menu), it seems to boot, or at least it booted now - trying to reboot | 15:35 |
| rbasak | hackeron: maybe also play with https://www.gnu.org/software/grub/manual/html_node/gfxpayload.html? | 15:35 |
| hackeron | rbasak: well, I donno what monitor may or may not be plugged in, so I don't want to hard code anything like that | 15:36 |
| hackeron | rbasak: and this worked fine in 14.10 - so it seems something in grub between 14.10 and 15.04 broke headless boot on Intel Atom and Intel Celeron machines like Shuttles and NUCs - not good | 15:37 |
| rbasak | hackeron: I was thinking of changing it to "text". | 15:39 |
| rbasak | hackeron: agreed it's not good. If it can be verified as a bug in grub or grub packaging I'm confident that it'll get fixed. Just need to pin it down. | 15:39 |
| hackeron | rbasak: will try - this server is in Moscow and I'm in London so coordinating is hard - I have around 70 servers, around 15 of them are experiencing this. Going to change the grub timeouts on them to 0 as a workaround, then assemble one locally and see if I can pin it down | 15:41 |
| rbasak | hackeron: thanks, I appreciate it. | 15:41 |
| hackeron | rbasak: no worries :) - also it seems setting timeouts to 0 isn't enough - it also needs GRUB_HIDDEN_TIMEOUT_QUIET=false | 15:46 |
| rbasak | hackeron: you are doing BIOS serial console redirection, are you? | 15:47 |
| rbasak | aren't doing | 15:48 |
| hackeron | what's that? - I'm not doing anythign special - I did do-release-upgrade to 15.04 and it stopped booting without a monitor | 15:48 |
| rbasak | hackeron: fairly common on headless boxes. The BIOS presents a serial device and redirects it to the monitor/keyboard. | 15:48 |
| rbasak | If that is enabled then I wonder if the BIOS is causing serial output to hang until the monitor is connected for example. | 15:49 |
| hackeron | rbasak: it's just an NUC | 15:49 |
| kickinz1 | hackeron, rbasak, I also experienced grub stuck at menu, when you run update-grub, there is a error message about GRUB_HIDDEN_TIMEOUT_QUIET and other related option not supported enabled at the same time anymore | 15:49 |
| rbasak | I've never handled a NUC physically, so I don't know much about them. | 15:49 |
| === matsubara_ is now known as matsubara | ||
| rbasak | I just don't follow why grub would care whether a monitor is plugged in. I didn't expect it to speak DDC. | 15:50 |
| kickinz1 | But I was having problem on making my system boot on a bcache rootfs, so I didn't take too much care of this. | 15:50 |
| rbasak | So I'm wondering if there's some kind of BIOS interaction going on here. | 15:50 |
| rbasak | And serial redirection is the first BIOS interaction I thought of that might be relevant. | 15:50 |
| kickinz1 | But why would the fact to connect the display make it start .... | 15:50 |
| kickinz1 | (was on 15.04) | 15:50 |
| hackeron | rbasak: ok, it seems GRUB_HIDDEN_TIMEOUT_QUIET=false isn't enough - need to remove splash and quiet also -- if the bios is doing that, why would removing splash and quiet and setting GRUB_HIDDEN_TIMEOUT_QUIET=false fix the issue? | 15:51 |
| hackeron | rbasak: and why wasn't it doing that in 14.10 or 14.04? | 15:53 |
| rbasak | hackeron: plenty of interactions change between releases. That doesn't necessarily make it a bug in a newer release. | 15:53 |
| kickinz1 | Hackeron; Do you have this warning when you update grub (translated from french): " Setting GRUB_TIMEOUT to a non-zero value if GRUB_HIDDEN_TIMEOUT is set is not possible anymore" | 15:54 |
| rbasak | hackeron: I am just speculating here. First let's figure out what's going on. Then we can talk about whether it's a bug and where and how it might be fixed. | 15:54 |
| hackeron | kickinz1: no warnings when running update-grub - I have all timeouts set to 0 | 15:54 |
| rbasak | hackeron: I'm suggesting that if you have BIOS serial redirection turned on, try turning it off to help pin down the bug. | 15:54 |
| kickinz1 | hackeron, OK, sorry fo noise, then. | 15:54 |
| rbasak | hackeron: I appreciate that's difficult in your situation. It's just a suggestion. | 15:55 |
| hackeron | rbasak: I very much doubt the NUC has this feature - this is just a tiny mini PC with an intel celeron CPU - also it's in Moscow -- I'll assemble one locally later today and check | 15:55 |
| rbasak | I have some cheap Atom based boards that do it. | 15:56 |
| rbasak | Don't NUCs have some basic management as well? | 15:56 |
| hackeron | rbasak: not that I know of.. - this also happened on a shuttle with an atom which just has defaults loaded in the bios - it's designed as a cheap desktop | 15:56 |
| hackeron | rbasak: this one: http://www.dabs.com/products/shuttle-slim-xs35v3l-black-barebone-system--atom-d2550-2-x-ddr3-sodi-949D.html?utm_source=google&utm_medium=ppc%20product%20search&utm_campaign=Computing%20-%20Desktops%20and%20Monitors%20-%20Desktop%20PCs&gclid=Cj0KEQjwvuuqBRDG95yR6tmfg9oBEiQAjE3RQNyYMACUOhj9Z06x02P_5OPcqAajBUV7pKRbVrcHVS8aAoM58P8HAQ | 15:57 |
| hackeron | actually it's a v2 with a VGA: http://us.shuttle.com/barebone/Models/XS35V2.html | 15:57 |
| teward | with an Apache web server is there a way to refresh the configurations (like how nginx can do `service nginx reload` and refresh the configurations without booting current connections or having the instance shut down)? | 16:05 |
| teward | (but close down the apache instance) | 16:05 |
| jrwren | teward: graceful | 16:06 |
| jrwren | teward: apachectl has a graceful option | 16:07 |
| teward | jrwren: OK, so that will not require a restart of the Apache process, and will gracefully apply the updated config for future connections without disrupting already connected connections? | 16:08 |
| teward | that's the problem (can't afford downtime!) | 16:08 |
| hxm | i have to say ubuntu 15 ruined my day | 16:08 |
| hxm | joking, but still true | 16:09 |
| jrwren | https://httpd.apache.org/docs/2.2/programs/apachectl.html yes | 16:09 |
| jrwren | " This differs from a normal restart in that currently open connections are not aborted." | 16:09 |
| teward | jrwren: ahhh, that's hyper important, thanks. | 16:09 |
| teward | i'm an nginx guy hence my asking :/ | 16:10 |
| teward | :P | 16:10 |
| teward | jrwren: stupid other question: what's the syntax to enforce SSL server side ordering in a config. | 16:10 |
| teward | in nginx it's ssl_prefer_server_ciphers, but no clue what it is in Apache | 16:10 |
| teward | (getting stuff dumped on me is blah) | 16:10 |
| jrwren | teward: SSLHonorCipherOrder on | 16:12 |
| jrwren | teward: https://httpd.apache.org/docs/current/mod/mod_ssl.html double check that its what you want. | 16:12 |
| hackeron | rbasak: an update, after some trial an error, it seems only this is required to make these servers boot: GRUB_CMDLINE_LINUX_DEFAULT="nomodeset" | 16:22 |
| hackeron | rbasak: so it seems like a kernel bug in that case? | 16:24 |
| hackeron | in fact this works too: GRUB_CMDLINE_LINUX_DEFAULT="nomodeset quite splash" | 16:25 |
| teward | jrwren: awesome, that's exactly what we needed. Thanks again :) | 16:25 |
| rbasak | hackeron: could be a kernel bug. Can you try the utopic kernel on vivid to pinpoint that? | 16:28 |
| hackeron | quiet* | 16:31 |
| hackeron | rbasak: sure, let me try - I believe I just change GRUB_DEFAULT=1 | 16:33 |
| rbasak | hackeron: yep - GRUB_DEFAULT. Though I'm never sure what 1 will do with older kernels being in a submenu now, so I do it by text: http://askubuntu.com/a/216420/7808 | 16:34 |
| rbasak | I match the exact text strings from the existing /boot/grub/grub.cfg. | 16:34 |
| rbasak | Then update-grub. | 16:34 |
| rbasak | '>' is for a submenu | 16:35 |
| hackeron | rbasak: for some reason it just booted the 15.04 kernel anyway and the sysadmin had to leave - will try to reproduce locally here | 16:57 |
| jKaideN | hey guys | 16:58 |
| jKaideN | anyone have experience with ILO2 (HP Integrated Lights-Out 2) ? | 16:58 |
| rbasak | hackeron: OK. I'm heading out now but will stay online and will check back later. If it does turn out to be a kernel issue, please file a bug - "kernel modesetting hangs headless machines until monitor is plugged in" or something. They'll probably want hardware details. | 16:58 |
| jKaideN | I'm trying to install my own ISO (OS) rather than using a cpanel with pre-set ISO's I'm not sure how exactly I can remotely install my own custom ISO | 16:59 |
| rbasak | hackeron: if you wanto dig further (we'd love it if you did) then https://wiki.ubuntu.com/Kernel/MainlineBuilds and https://wiki.ubuntu.com/Kernel/KernelBisection will hopefully help pin it down to a single commit. | 17:00 |
| jKaideN | in the ILO2 web management I have this thing called "Virtual Media Applet" But i just get java error | 17:02 |
| hackeron | rbasak: thank you, I'm just buiding a new NUC here to try to reproduce the problem first :) | 17:04 |
| === Lcawte is now known as Lcawte|Away | ||
| === Lcawte|Away is now known as Lcawte | ||
| === pgraner is now known as pgraner-food | ||
| pmatulis | jKaideN: using windows/IE on the client side? if so, that should work | 17:47 |
| jKaideN | yeah i am not working though | 17:47 |
| jKaideN | Could not initialize class com.sun.deploy.net.protocol.https.Handler$Intializer | 17:48 |
| jKaideN | Ah disabled TLS 1.1 + 1.2 on Java | 17:50 |
| jKaideN | it's working now | 17:50 |
| jKaideN | Ubuntu Server 15.04 or Ubuntu Server 14.04.2 LTS ?? | 17:53 |
| hackeron | rbasak: argh! - I have the same NUC here but a different revision and cannot reproduce the problem here, but 15 of 60 servers have this problem now and they are in client locations so can't really ask them to keep rebooting while I experiment :( - going to try the workaroud and see if the problem goes away for them all | 17:55 |
| rbasak | hackeron: a higher or lower revision? I wonder if it was a BIOS bug that got fixed? | 18:07 |
| hackeron | rbasak: same bios version, different hardware revision | 18:07 |
| rbasak | OK, I guess not then. | 18:08 |
| hzut | Hi - Please, let me know a good successor for emule? | 18:19 |
| jKaideN | Hey guys anyone here I got a quick question regarding "Virtual Media" | 18:21 |
| jKaideN | now, i'm installing Ubuntu Server on my remote dedicated server | 18:21 |
| jKaideN | I've got access to KVM switch which has things like remote console to server and this thing called "Virtual Media" | 18:21 |
| jKaideN | i've located the virtual media to the ISO i downloaded (ubuntu server) on my client local machine | 18:22 |
| jKaideN | Does this upload the iso ? because the iso is around 600 MB and my upload speed is really slow | 18:22 |
| jKaideN | was wondering how is the server reading the file from my local PC without me having to upload it ?? | 18:23 |
| hackeron | rbasak: the plot thickens - I managed to reproduce the problem, but only when doing do-release-upgrade from 14.04 -> 14.10 -> 15.04 -- if skipping 14.04, there are no boot problems - how bizare :/ | 18:36 |
| jrwren | hackeron: this is the grub thing, right? can you diff -r /boot between a system exhibiting the problem and another not? | 18:44 |
| === pgraner-food is now known as pgraner | ||
| hackeron | jrwren: will do, just helping a client get a system back up and running - thankfully the workaround fixes the immediate issue | 18:59 |
| pmatulis | jKaideN: yes, will be slow | 19:09 |
| jKaideN | pmatulis, yep i'm setting up right now, um i've got 2 network interfances i've just setup eth0 as primary | 19:10 |
| jKaideN | how would i know if thats the correct one ? | 19:10 |
| jKaideN | and im stuck at "partitioning disks" | 19:10 |
| pmatulis | stuck how? | 19:11 |
| jKaideN | Guided - Use entire disk ? | 19:11 |
| jKaideN | so many options | 19:11 |
| jKaideN | or should i be doing manual for this | 19:11 |
| jKaideN | server has 1 500GB HDD | 19:11 |
| pmatulis | jKaideN: depends what you want in the end | 19:11 |
| pmatulis | jKaideN: do you need any special partitioning? | 19:12 |
| jKaideN | nope | 19:12 |
| jKaideN | I don't | 19:12 |
| pmatulis | jKaideN: choose 'guided + LVM' | 19:12 |
| jKaideN | Guided - use entire disk and set up LVM ? | 19:12 |
| pmatulis | yep | 19:13 |
| pmatulis | then accpet the next few questions | 19:13 |
| pmatulis | (yes) | 19:13 |
| jKaideN | it's asking for volume group to use for guided partitioning: | 19:14 |
| jKaideN | it automatically says "499.8 GB" leave it as it is/ | 19:14 |
| pmatulis | yes | 19:14 |
| jKaideN | http://i.imgur.com/heR2NoS.png | 19:15 |
| jKaideN | is that all good ? | 19:15 |
| pmatulis | yes | 19:16 |
| jKaideN | upload speed : 47.37 KB/s :( | 19:26 |
| jKaideN | #Australia | 19:26 |
| sarnold | ow :/ | 19:26 |
| pmatulis | go australia, those packets are taking the scenic route around corrals and whateva | 19:28 |
| hzut | Hi - Please, let me know a good successor for emule? | 19:31 |
| jKaideN | lol and my download is 100 MB/s rofl | 19:32 |
| jKaideN | for the nameserver i just set it to 8.8.8.8 | 19:40 |
| jKaideN | thats fine ? | 19:40 |
| === NomadJim_ is now known as NomadJim | ||
| pmatulis | that's local-specific | 19:41 |
| jKaideN | ? | 19:41 |
| Walex2 | hzut: 'amule' | 20:04 |
| jKaideN | how should i manage upgrades on the system? which is preferred? | 20:51 |
| jKaideN | no auto updates | install security updates auto | manage system with landscape | 20:51 |
| bekks | jKaideN: The way that fits your requirements. | 20:57 |
| CrustY__ | Hi everyone | 20:57 |
| CrustY__ | I've faced this issue https://bugs.launchpad.net/ubuntu/+source/policykit-1/+bug/1447654 | 20:57 |
| CrustY__ | and tried to install polkit like in workaround described in comment | 20:58 |
| jKaideN | I just went with no auto updates i'll porbably want to manage them myself | 20:58 |
| CrustY__ | but apt-get install polkit-1 hangs on installing udev | 20:58 |
| CrustY__ | how can I fix that? | 20:58 |
| jKaideN | when you first install ubuntu server is it essential to do 'apt-get full-upgrade' ? | 21:01 |
| bekks | jKaideN: apt-get dist-upgrade | 21:01 |
| CrustY__ | I use it on cloud | 21:01 |
| CrustY__ | to be clearer: it's droplet in digitalocean | 21:02 |
| bekks | jKaideN: full-upgrade isnt even mentioned in the man page. | 21:02 |
| jKaideN | bekks, oh i see, is the dist-upgrade any different from normal upgrade | 21:02 |
| jKaideN | bekks, in other words why should i be doing dist-upgrade opposed to normal upgrade | 21:03 |
| bekks | jKaideN: This article explains it: http://askubuntu.com/questions/81585/what-is-dist-upgrade-and-why-does-it-upgrade-more-than-upgrade | 21:03 |
| CrustY__ | any help?:( | 21:10 |
| jKaideN | do i install GRUB boot loader? | 21:17 |
| sarnold | CrustY__: perhaps add a comment to the bug with the information pitti asked for; he'll probably be online in seven or eight hours and might be able to work on it then | 21:57 |
| pmatulis | jKaideN: 'apt full-upgrade' will work | 22:15 |
| pmatulis | jKaideN: yes, install GRUB | 22:16 |
| jKaideN | yep done | 22:16 |
| sarnold | heh, I've never seen full-upgrade before. | 22:16 |
| bekks | pmatulis: How comes it isnt documented in the man page? | 22:16 |
| jKaideN | installing iptables-persistent now | 22:16 |
| pmatulis | apt not apt-get | 22:16 |
| bekks | ah, oh, ok. TIL. :) | 22:17 |
| pmatulis | hopefully to have completion soon | 22:17 |
| bekks | :) | 22:17 |
| jKaideN | btw im using ufw will that conflict with iptables ? | 22:18 |
| jKaideN | like when i allow something on ufw but accidentally deny in iptables which takes highest priority ? | 22:18 |
| sarnold | jKaideN: ufw writes iptables rules for you | 22:18 |
| bekks | ufw uses iptables, but you're better of using either ufw for administering iptables, or dont use ufw. | 22:18 |
| sarnold | jKaideN: I'd recommend picking one; hand-writing rules or using ufw, and stick with it... | 22:18 |
| jKaideN | ok since i'm not going to be using much rules everything will be kept simple i'll just go with ufw | 22:19 |
| jKaideN | I have | 22:27 |
| jKaideN | -P INPUT DROP | 22:27 |
| jKaideN | -P FORWARD DROP | 22:27 |
| jKaideN | is that normal | 22:27 |
| jKaideN | is there a way to update python and making sure that two different versions are non-existing? | 22:34 |
| sarnold | jKaideN: what are you trying to do? | 22:34 |
| sarnold | jKaideN: hint: it's best to pretend python3 and python2 are completely different languages | 22:34 |
| jKaideN | update current python from 2.7.6 to 2.7.9 | 22:35 |
| jKaideN | yes | 22:35 |
| sarnold | jKaideN: the easiest way to do that is to upgrade to vivid. | 22:35 |
| sarnold | jKaideN: alternatively you could compile your own python fro msource for whatever requires 2.7.9. | 22:36 |
| jKaideN | things that use python doesn't require 2.7.9 but I'd like to update python for security reasons as well, rather not keep them outdated | 22:38 |
| jKaideN | example: fail2ban requires python >= v2.6 (or >= 3.2) | 22:40 |
| sarnold | 2,7,6 fits that.. | 22:40 |
| sarnold | there's not much outstanding for python2.7: http://people.canonical.com/~ubuntu-security/cve/pkg/python2.7.html | 22:40 |
| sarnold | we've rated all those low priority issues | 22:41 |
| jKaideN | i should be fine with 2.7.6 ? | 22:42 |
| jKaideN | i'm using ufw does this save the table after reboot | 22:50 |
| jKaideN | or am i required to install iptables-persistent as well | 22:50 |
| jKaideN | also in motd i receive "7 updates are security updates. | 22:53 |
| jKaideN | " How do i apply these updates ? through apt-get upgrade ? | 22:53 |
| === Guest38176 is now known as mgagne | ||
| bekks | jKaideN: Just use apt-get dist-upgrade, instead of aptget upgrade, always. | 23:06 |
| jKaideN | also on syslog i got something like this: | 23:07 |
| jKaideN | [UFW BLOCK] IN=eth0 OUT= MAC=00:19:.... SRC=128.8.x.x DST=<my server ip> LEN=60 TOS=0x00 PREC=0x00 TTL=47 ID=12690 DF PROTO=TCP SPT=50848 DPT=8333 WINDOW=14600 RES=0x00 SYN URGP=0 | 23:08 |
| jKaideN | there's a couple of htem | 23:09 |
| jKaideN | lol wtf that ip is dreyfus.umiacs.umd.edu | 23:12 |
| === Lcawte is now known as Lcawte|Away | ||
Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!
