lwizardl | hello | 04:23 |
---|---|---|
ramsay_za | morning is there any way to setup mdadm raid arrays when juju bootstraps a server? | 06:59 |
elemay2 | hi, im on ubuntu server 12.04 x64, i have 2 nics (both e1000 / Intel Pro 1000). These are configured as bond0, one of them is only getting 100mbps link. i changed the card already with another one and also changed the nw-cable, i also switched the ports on the switch. any hints for this? | 08:12 |
=== smb` is now known as smb | ||
sveinse | I'm running 12.04 amd64 server (kernel 3.2.0-37-generic #58) on a VMware vCenter Server (5.1.0). I'm having regular problems with the kernel crashing. Invalid opcode 0000 in highmem.h:194. Ubuntu runs the latest VMWareTools. Any ideas? | 08:33 |
sveinse | Natty on the same datacenter does not show this behaviour | 08:34 |
=== sweettea is now known as Guest30093 | ||
vedic | How to disable single user mode login? or is there any way to protect colocated servers at remote place? (I do need server restart though) | 08:41 |
cnf | hi | 08:47 |
cnf | when using a preseed installation, is there a way to get the installer to ask confirmation for prefilled values? | 08:47 |
rbasak | cnf: you can avoid setting priority on the kernel commandline. Or set it to low for even more questions | 08:50 |
cnf | rbasak: how does that get me prefilled questions? | 08:50 |
cnf | i want no questions, except for network configuration | 08:51 |
cnf | and i want those fields prefilled with defaul values | 08:51 |
rbasak | Oh, I see. I'm not sure you can do that. | 08:51 |
rbasak | Questions you are asked are based on their priority | 08:51 |
cnf | hmm | 08:51 |
rbasak | Preseeding sets defaults in advance. I believe the two mechanisms are entirely independent | 08:52 |
cnf | i can get it to prompt for network settings | 08:52 |
cnf | but i would like to prepopulate for example our dns servers | 08:52 |
rbasak | Maybe there's another mechanism I'm unaware of | 08:52 |
rbasak | Why not just use DHCP? | 08:52 |
cnf | uhm, because there is no dhcp on our network | 08:52 |
cnf | at all | 08:53 |
cnf | anywhere | 08:53 |
rbasak | Put one in then :) | 08:53 |
cnf | why would we do that? | 08:53 |
rbasak | So that you don't have to input network settings | 08:53 |
cnf | we'd have to run maybe 50 dhcp servers | 08:53 |
cnf | not much of a solution :P | 08:54 |
rbasak | Crazy | 08:54 |
cnf | rbasak: probably about 200 physical machines (not counting customer machines) with in all 300 or 400 OSes running on 50ih vlans? | 08:54 |
rbasak | At that scale you should have everything fully automated then | 08:55 |
cnf | legacy is a horrible thing :P | 08:55 |
rbasak | Use a provisioning vlan with one dhcp server | 08:55 |
cnf | which entails switching around vlans left and right | 08:56 |
cnf | like i said, legacy is a horrible thing | 08:56 |
vedic | How to disable single user login to improve security of the server? What is the best practice? | 09:35 |
agu10^ | I am unable to understand how to set up virtual hosts properly. Tutorials send me to edit DIFFERENT config files. This inconsistency troubles me. am I not supposed to have one file per virtual host at sites-available ?? tutorials make me edit httpd.conf, or add a new file "virtual.conf", or stuff like that. But i believe neither is the right approach. any clues? I have added a file at /etc/apache2/sites-available/potentialstudio.com . | 09:39 |
agu10^ | .. but it's not doing anything | 09:39 |
cnf | rbasak: from #ubuntu-installer: | 09:44 |
cnf | <cjwatson> d-i name/of/question string value | 09:44 |
cnf | <cjwatson> d-i name/of/question seen false | 09:44 |
cnf | <cjwatson> https://help.ubuntu.com/12.04/installation-guide/i386/preseed-advanced.html#preseed-seenflag | 09:44 |
cnf | for future reference | 09:45 |
rbasak | Ah, handy | 09:45 |
rbasak | Thanks | 09:45 |
cnf | yeah, indeed | 09:45 |
jpds | vedic: What are you trying to do? | 09:57 |
jpds | vedic: Disable the recovery mode? | 09:57 |
vedic | jpds: Server is colocated and placed at remote location. Just thought to increase its security if somebody try logging in as single user mode | 10:00 |
jpds | vedic: If they have physical access, there's not much you can do. | 10:00 |
vedic | jpds: yea, live cd is always an option. | 10:01 |
jpds | vedic: LiveCD, ripping the hard drives out. | 10:01 |
jpds | vedic: All I can really suggest is putting non-system files on an encrypted partition. | 10:04 |
jpds | But then every time the server boots, you'll have to manually mount that partition. | 10:04 |
vedic | jpds: I think that would stop running programs reading files. I have my codes and data text files in home dir and encrypting that would cause programs to not function properly | 10:05 |
jpds | vedic: Yeah, I'm thinking non-home dir encryption. | 10:06 |
hackeron | I'm trying to install ubuntu on a blank hard-drive connecting to an exiting ubuntu server. I have most of the progress figured out - but I cannot get it to boot. I just see a white blinking cursor and the grub menu never shows up. I think my partitioning code must be wrong: http://pastie.org/6173525 -- any ideas what is wrong with that? | 11:23 |
cfhowlett | hackeron, probably a grub configuration issue, i.e. grub has to be directed to the bootable media | 11:24 |
freesbie | hackeron: have you done a grub-install on the new disk ? | 11:25 |
hackeron | freesbie: yes, I did grub-install $DRIVE - it said installed without issues | 11:26 |
hackeron | freesbie: if I change the label from gpt to msdos - instead of a blinking white cursor, I get a reboot loop | 11:27 |
hackeron | cfhowlett: yes, this is what I'm doing with parted $DRIVE -- set 1 bios_grub on | 11:28 |
freesbie | hackeron: you may have to give it --boot-directory=DIR aswell | 11:52 |
hackeron | freesbie: but it's /boot which is the default :/ | 11:58 |
agu10^ | I can't get apache2 to start! root@domU-12-31-39-10-4A-43:/etc/apache2# service apache2 start | 11:59 |
agu10^ | * Starting web server apache2 Action 'start' failed. | 11:59 |
agu10^ | The Apache error log may have more information. | 11:59 |
greppy | agu10^: what do the logs tell you? | 12:00 |
agu10^ | nothing at all | 12:00 |
agu10^ | the last log was hours ago | 12:01 |
agu10^ | when it still worked | 12:01 |
greppy | so there is nothing in /var/log/apache2/error.log? | 12:01 |
agu10^ | nothing new. and nothing related to this failure | 12:01 |
agu10^ | last log there is from 2 hours ago | 12:01 |
agu10^ | a GET request | 12:01 |
greppy | is that where it is still configured to log to? | 12:01 |
agu10^ | but now it's not even STARTING | 12:01 |
agu10^ | yes, as far as i know, yes | 12:02 |
agu10^ | i didn't change it | 12:02 |
greppy | agu10^: what about sudo apache2ctl configtest | 12:04 |
greppy | ? | 12:04 |
agu10^ | Syntax OK | 12:04 |
greppy | got me then. | 12:04 |
greppy | I'd double check that the logs are still going to where you expect | 12:05 |
agu10^ | how do i know? | 12:06 |
greppy | look in the configuration files in /etc/apache2 | 12:06 |
bitnumus | Hi, im trying to install ubuntu-server, the installation failed at loading the CD-ROM stage, any way to find out why? | 12:06 |
agu10^ | greppy, i did | 12:07 |
cfhowlett | bitnumus, any error messages? also did you md5sum the iso? | 12:08 |
agu10^ | ErrorLog ${APACHE_LOG_DIR}/error.log | 12:08 |
bitnumus | i did md5sum the ISO yea | 12:08 |
bitnumus | it failed 4 times, i tried formatting the USB to fat32 etc | 12:08 |
gregory58 | Hello, I am following a Railscast(http://railscasts.com/episodes/335-deploying-to-a-vps) to deploy to DO, but I am getting a "bash: vim: command not found" error when I input "vim ~/.bashrc " | 12:09 |
cfhowlett | bitnumus, bootable cd/dvd? or usb? | 12:09 |
bitnumus | cfhowlett, USB | 12:09 |
bitnumus | using 'unetbootin' | 12:09 |
cfhowlett | bitnumus, easy check: different usb. they do get squirrely ... | 12:09 |
bitnumus | hmm, not sure i have another one handy | 12:10 |
bitnumus | worth 00'ing it or something? | 12:10 |
cfhowlett | bitnumus, just format as normal I'd say | 12:11 |
bitnumus | yea i did that using gparted | 12:11 |
bitnumus | cfhowlett, i ran 'check disk integrity' and it fails on a file saying 'file failed the Md5 checksum verifitcation' | 12:13 |
bitnumus | Is there an alternative to unetbootin i can try, to rule that out ? | 12:13 |
hackeron | freesbie: I get this: # grub-install --boot-directory=/boot /dev/sdi | 12:13 |
hackeron | Installation finished. No error reported. | 12:13 |
hackeron | but when I try to boot, I just see a white blinking cursor and the grub menu is never shown - I think it is the partition table to blame | 12:14 |
cfhowlett | bitnumus, this is your downloaded iso? if it failed download again. I'd use the torrent for best results... | 12:16 |
bitnumus | cfhowlett, i downloaded it from releases.ubuntu.com | 12:17 |
bitnumus | and the md5sum of the download is correct | 12:17 |
cfhowlett | bitnumus, so on installation it reads an error? the bootable usb is fragged. do it again - perhaps with startupdiskcreator? | 12:18 |
bitnumus | im trying startup disk creator now | 12:18 |
bitnumus | i ran 'check disk integrity' and it fails on a file saying 'file failed the Md5 checksum verifitcation' | 12:18 |
bitnumus | it fails on loading files from CD-ROM | 12:18 |
bitnumus | it boots fine | 12:19 |
cfhowlett | bitnumus, you're reading inconsistent feedback ... clearly there's an error here. either hardware or software fail... | 12:20 |
gregory58 | Anyone know how to install vim on a Ubuntu VPS like Digital Ocean....I'm getting a vim: command not found error | 12:20 |
freesbie | hackeron: maybe it is, I remember having some problems with GPT and grub, had to make a bit of free unpartitioned space to make it work. but that was on virtual machines, afair I initiated grub from within a chroot on the disk-image | 12:20 |
cfhowlett | bitnumus, option? install the mini.iso and download then add the packages you need via internet... | 12:21 |
cfhowlett | and don't use CD-rom? | 12:21 |
bitnumus | cfhowlett, ill try this startup disk creator and see what happens | 12:21 |
bitnumus | im not using a CD-ROM | 12:21 |
hackeron | freesbie: yeh, that's pretty much what I'm doing except not on a virtual machine - I'm doing chroot /tmp/target and then running the grub-install - I tried to leave the free space in my partitioning: http://pastie.org/6173525 - but it just wont boot :/ | 12:22 |
bitnumus | thats what the installer calls it | 12:22 |
mah454 | Hello | 12:34 |
mah454 | I installed mysql in ubuntu 12.10 | 12:34 |
mah454 | but have not log in /var/log for mysql ! | 12:34 |
mah454 | How can enable mysql log ? | 12:35 |
andol | mah454: /etc/mysql/my.cnf | 12:36 |
mah454 | andol: what line must change ? | 12:36 |
andol | mah454: That depends what kind of logging you want. Tried searching for the word log in the file? | 12:37 |
mah454 | andol: mysql authentication (User Login or Logout) | 12:39 |
andol | Not sure about that particular case. | 12:41 |
agu10^ | I try to create a new "default" website for apache2 but with the name "wildcard" | 12:45 |
agu10^ | is it impossible to create it under a name other than "default" ??? | 12:45 |
=== yofel_ is now known as yofel | ||
Daviey | zul: thanks for replying to the openstack list. | 12:52 |
zul | Daviey: no worries | 12:52 |
jamespage | zul: http://people.canonical.com/~jamespage/ca-updates/ review please | 13:07 |
jamespage | :-) | 13:07 |
zul | awesome-o +1 | 13:07 |
Daviey | that was the fastest review ever. :) | 13:11 |
=== wedgwood_away is now known as wedgwood | ||
=== Sophism is now known as Guest37491 | ||
jamespage | ivoks, think I figured out some of those glance errors from the other day - haproxy had very low server/client timeouts | 14:53 |
jamespage | I suspect nothing happened for a bit so it killed the connections | 14:53 |
=== unix is now known as megha | ||
=== Guest37491 is now known as ckelly | ||
zul | smoser: ping | 16:19 |
smoser | hey | 16:19 |
zul | smoser: this is the dirty fix to get it going again http://paste.ubuntu.com/1658525/ | 16:20 |
zul | smoser: im coming up with a bit more elegant solution ;) | 16:20 |
smoser | zul, is the issue that they're trying to use libvirt's mounting of disks ? as opposed to having nova set that up and providing libvirt with a path ? | 16:22 |
sliddjur | How do I list a packages dependencies? ubuntu server 10 | 16:23 |
zul | smoser: thats the eventual goal i think | 16:23 |
zul | smoser: but the disk doesnt get copied to the instances directory right now because mapping is an empty arrary because blockinfo.py doesnt understand lxc block devices (hint: there isnt any) | 16:25 |
smoser | sliddjur, apt-get install --dry-run <package> | 16:27 |
smoser | will give you a list of what it would install for you | 16:27 |
smoser | apt-cache show <package> | 16:27 |
smoser | will show you its immediate dependencies | 16:27 |
sliddjur | smoser: apt-cache show package | grep Depends worked out well =) | 16:28 |
zul | smoser: but one of the problems i see with lxc block devices is that the domain gets created before the images get created and keys are injected | 16:37 |
smoser | well, we dont really care about injected keys. | 16:38 |
smoser | but we will probably need files to come from somewhere | 16:38 |
Eagleman | Is there being worked on a backport for libvirt 1.01+ ? | 17:03 |
ivoks | jamespage: ive tried without haproxy; same thing :-/ | 17:06 |
iqualfragile | the ubuntu serverguide suggests using only ldap for authentification, is it safe to do so? (instead of using kerberos) | 17:09 |
iqualfragile | well, i guess not | 17:11 |
iqualfragile | but i like the idea of kerberos, will use it anyways | 17:11 |
=== JanC_ is now known as JanC | ||
jamespage | ivoks, same problem with horizon right? | 17:26 |
=== chuck_ is now known as zul | ||
keithzg_ | Rather random question, I'm using s3cmd to sync some backups and on every file I get "WARNING: path/to/file: Owner groupname not known. Storing GID-3342 instead." I'm inexperienced enough with S3 that I really don't know what that implies (otherwise the backups seem fine). | 18:28 |
NewNewb | Good evening/day/morning! Ive got issues with a game server running on ubuntu through wine. Problem is, I cannot connect to my game server - and I have no clue how to do error searching. Ideas?` | 18:30 |
sarnold | NewNewb: is the game server hosted on a different network than the clients trying to connect? | 18:31 |
NewNewb | Its on my LAN. and connecting through local IP, nor external works(for my friends). The server itself shows up in the game's server browser | 18:32 |
sarnold | NewNewb: are there any log entries in /var/log/ that look related? | 18:33 |
sarnold | NewNewb: does wine preserve enough of windows to have the windows event log that might also have information from the game server? | 18:33 |
NewNewb | I tried syslog, nothing at all. And I tried messages, also nothing, it just says "--MARK--" allt he time | 18:33 |
sarnold | NewNewb: windows allows any process to bind to "low ports", < 1024 -- Linux does not. a process must be privileged to bind to a low port... | 18:34 |
sarnold | that --MARK-- will become reassuring eventually :) | 18:34 |
NewNewb | ports of relevance here are 7778, 7777, and 27015 | 18:34 |
Eagleman | Is there being worked on a backport for libvirt 1.01+ ? | 18:35 |
NewNewb | 7777 being the port external connections come through. 7778 lists the server (i think) and 27015 is a steam port... no clue what that one does :) | 18:35 |
sarnold | NewNewb: darn, there goes another potentially easy answer :) | 18:35 |
NewNewb | haha, ive been hitting up forums and google for two weeks now. I just suck at linux - and dont know how to do proper error checking :( | 18:36 |
aaa801 | im running a java server (minecraft) in a screen session, every soo often the server goes down, when i open the screen session i find 19:28:19 [INFO] Unloading Player: Gremlin503 | 19:42 |
aaa801 | >Killed | 19:42 |
aaa801 | root@ns3273339:/home/root/Voltz# | 19:42 |
aaa801 | Does that mean something on the box is killing the process ? | 19:42 |
sarnold | is it always gremlin? :) | 19:42 |
aaa801 | no | 19:43 |
aaa801 | Out of memory: Kill process 14432 (java) score 859 or sacrifice child | 19:43 |
aaa801 | Killed process 14432 (java) total-vm:38229524kB, anon-rss:30098300kB, file-rss:0kB | 19:43 |
aaa801 | son of a.. | 19:43 |
aaa801 | Welp i think i juts found the issue | 19:43 |
sarnold | aha, you anticipated my next question :) | 19:43 |
* aaa801 walks off in shame | 19:44 | |
sarnold | thgh, uh, one does wonder what java intended to do with 36 gigabytes of memory. | 19:44 |
aaa801 | i only had it assigned 30gb | 19:45 |
aaa801 | =/ | 19:45 |
aaa801 | how it managed to even get hold of 36gb is odd | 19:45 |
aaa801 | im giving it 26g now | 19:46 |
sarnold | aaa801: sorry.... it had allocated 36 gigs, but used 28 or so... | 19:47 |
sarnold | is that expected for minecraft? o_O | 19:47 |
aaa801 | for this server ye | 19:47 |
aaa801 | 100 slots | 19:47 |
aaa801 | mc is terrible for memory management + java is too | 19:47 |
sarnold | damn :) | 19:47 |
aaa801 | still not sure it had allocated 36 gig | 19:48 |
aaa801 | it was started with a max of 30g defined =/ | 19:48 |
sarnold | aaa801: well, virtual memory that was allocated by potentially never actually used... | 19:50 |
sarnold | well, you can see the RSS there, it wasn't used.. :) | 19:50 |
aaa801 | then whyd it kill the process | 19:50 |
aaa801 | >_> | 19:50 |
sarnold | aaa801: how much RAM does the machine have? | 19:50 |
aaa801 | 32gb | 19:50 |
sarnold | aaa801: does it run anything else that might have taken up 4-odd gigs? | 19:51 |
aaa801 | shouldnt do, its got two ssd's in raid | 19:51 |
aaa801 | software raid | 19:51 |
sarnold | aaa801: hrm. the kernel should normally only start killing things once it is in trouble. | 19:52 |
aaa801 | let me open up top | 19:52 |
sarnold | aaa801: you could add some swap files ... dd to create them, mkswap to format them, then swapon to add them. | 19:52 |
aaa801 | http://i.imgur.com/MabnN78.png | 19:53 |
aaa801 | how is it managing to take up 31.6g | 19:53 |
aaa801 | when its limited to 26g | 19:53 |
aaa801 | -.- | 19:53 |
aaa801 | sarnold, any ideas? | 19:53 |
sarnold | aaa801: well, right now it is only taking up ten gigs... | 19:54 |
sarnold | aaa801: oooh | 19:54 |
aaa801 | oooh? | 19:54 |
sarnold | aaa801: hrm, maybe 'oooh' in a bit. | 19:55 |
aaa801 | ._. | 19:55 |
sarnold | aaa801: perhaps the 31 vs 26 is just 1000 vs 1024 nonsense? | 19:55 |
aaa801 | still shouldnt be that high when its on 26G | 19:55 |
sarnold | no, 28g at the most if that were it... | 19:55 |
sarnold | aaa801: okay, back to 'oooh' | 19:55 |
sarnold | aaa801: if you don't want to add some swap space, you could tweak /proc/sys/vm/overcommit_memory or /proc/sys/vm/overcommit_ratio to force memory allocations by java to fail when the machine really has run out of memory | 19:56 |
aaa801 | swap space and ssds dont tend to mix too well | 19:56 |
sarnold | aaa801: yeah, I know what you mean. but it might not be used too frequently.. | 19:57 |
sarnold | (you' have to chech with e.g. vmstat 1 once it is running again to find out) | 19:57 |
sarnold | aaa801: you can find details on the proc files in the proc(5) manpage | 19:58 |
aaa801 | kk | 19:58 |
sarnold | (sorry I left that out earlier. sigh :) | 19:58 |
subman_ | I have setup Munin according to the instructions found in the Ubuntu Server Guide and it works for the localhost only. I don't seem to be able to see the data from the second server (munin-node). Any ideas why this would be? | 20:03 |
zul | smoser: https://review.openstack.org/#/c/22065/ | 20:25 |
yeats | subman_: check your syslog on the remote host to see if there are any messages from munin | 20:25 |
yeats | I would also see if there's a way to run munin's daemon in a debug mode | 20:26 |
smoser | zul could you put the trace in the bug ? | 20:27 |
smoser | the one that is there now. | 20:27 |
smoser | just for reference. | 20:27 |
zul | smoser: sure | 20:27 |
smoser | https://review.openstack.org/#/c/22065/1/nova/virt/libvirt/blockinfo.py | 20:29 |
smoser | it might be better to return None in line 119 there. | 20:29 |
smoser | or maybe 'LXC'. the "" i'm afraid will just confuse things at some point. | 20:30 |
subman_ | yeats, thanks, it was a hostname issue. | 20:31 |
smoser | zul, ^ | 20:36 |
zul | smoser: good idea | 20:37 |
zul | smoser: lets see what other people say as well I might be tied to the stake as well | 20:37 |
smoser | i should put my comment in a review | 20:38 |
ivoks | jamespage: yes | 21:22 |
=== wedgwood is now known as wedgwood_away | ||
=== wedgwood_away is now known as wedgwood | ||
iqualfragile1 | is there some way i (well not i but someone who does not now too much about computers) can manage users wich reside inside ldap and kerberos? | 21:57 |
RyanLMul | Anyone run across an instance where 12.04 server stops communicating with DNS on another 12.04 intermittently? | 22:18 |
RyanLMul | Hi there. Anyone on here? | 22:28 |
sarnold | RyanLMul: a few hundred .. I guess no one reading has seen what you've seen. | 22:29 |
aaa801 | Remember children 99.99999% of people in the irc channel are idle | 22:37 |
=== JoakimKoed is now known as vooze |
Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!