[00:53] hi guys any idea how do i install lame apps-- coz when i tried with my ubuntu server 10.04LTS -http://pastebin.com/gv8FfYBv [00:53] any idea guys..? [02:34] hi guys [02:36] trying to log onto the amazon cloud to no avail...I have looked at https://10.cloud.ubuntu.com/faq/ but where do get the cloud VM ? [03:10] hey guys, using ubuntu server 11.04, trying to setup isc-dhcp4 and I keep getting a segment fault [03:12] anyone around? [04:31] hey everyone, any reason why /usr/sbin/dhcpd uses a lot of cpu and mem? after killing the process server load droped from 1.56 to 0.27 and mem usage went from 76% to 0% [05:03] What would be the recommended filesystem these days for a 6TB RAID5? [05:03] this is for a home fileserver, so lots of big video files, small music files, reasonable amount of varied simultaneous access [05:04] but also preferably something that doesn't just truncate all open files on power loss [05:04] (I actually had that happen with a FS once =( ) [05:05] also, i'm running 10.04, that might have some effect on the maturity of what's available in my kernel... [05:21] hello, I've run into a problem where /var/lib/ureadahead/debugfs is taking up all my system space as reported by df to the point that I can't upgrade or install new packages... I've read it is just a reporting error... but how do I get rid of it. [05:21] ref: http://ubuntuforums.org/showthread.php?t=1350785 [05:21] http://ubuntuforums.org/showthread.php?t=1670411 [05:24] Is this the appropriate solution? http://ubuntuguide.net/howto-fix-ureadahead-problem-after-upgrading-to-ubuntu-10-04 [05:32] I am running a bot, and it keeps telling me that it can't load a module [05:32] I checked the module and everything seems to be fine with it [05:32] and I have the appropriate plgins installed.l [06:01] New bug: #779336 in mysql-5.1 (main) "package mysql-server-5.1 5.1.54-1ubuntu4 failed to install/upgrade: subprocess installed post-installation script returned error exit status 1" [Undecided,New] https://launchpad.net/bugs/779336 [08:11] New bug: #779356 in ipsec-tools (main) "racoon and iPhone workaround" [Undecided,New] https://launchpad.net/bugs/779356 [08:32] New bug: #779362 in php5 (main) "package libapache2-mod-php5 5.3.5-1ubuntu7.2 failed to install/upgrade: subprocess installed post-installation script returned error exit status 128" [Undecided,New] https://launchpad.net/bugs/779362 === med_out is now known as medberry [10:12] for the record Bacta is banned [10:12] nothing has changed with him [10:19] O_O [10:21] New bug: #779391 in exim4 (main) "CVE-2011-1764: format string vulnerability" [Undecided,New] https://launchpad.net/bugs/779391 [11:29] Hey! I'm running an ubuntu lucid webserver configuration and switched from mod_php to mod_fcgid to benefit from apache's mpm_worker. The server response is better than under mod_php but under load it performs worse and my error.log keeps being flooded with notices like this: [Thu May 05 12:16:37 2011] [notice] mod_fcgid: /aegir/drupal-7/index.php total process count 15 >= 15, skip the spawn request. My mod_fcgid configura [11:29] http://pastebin.com/6H1HkyLh . I use apc as an opcode cache & memcache for object caching. Under load the server doesn't start swap... Anyone an idea how to solve this???? [14:36] What would be the recommended filesystem these days for a 6TB RAID5? [14:36] this is for a home fileserver, so lots of big video files, small music files, reasonable amount of varied simultaneous access [14:37] but also preferably something that doesn't just truncate all open files on power loss [14:37] (I actually had that happen with a FS once =( ) [14:37] also, i'm running 10.04, that might have some effect on the maturity of what's available in my kernel... === mns` is now known as Guest31355 [14:50] Alan: zfs? ;) [14:50] Alan: really, just use ext4 [14:51] * RoyK uses ZFS for such storage, but not on linux [14:51] I've never seen linux truncate files on power loss [14:52] you may lose parts of a file's contents if you get powerloss while writing, but that happens to all filesystems unless you have battery backed cache (or using an SLOG SSD with ZFS) [14:57] yeah, with modern large disks, invisible data corruption happens [14:58] next time I expand my array, I plan on switching to zfs for that reason [15:01] my current one uses ext3, which has been good to me [15:05] RoyK: it was a few years ago, using ZFS, every open file got truncated to 0 bytes when i accidentally hit the reset button once [15:06] RoyK: is ext4 properly stable in the 10.04 kernel? [15:06] yes [15:07] it's the default kernel in 10.04LTS [15:09] by kernel do you mean FS? [15:10] I can't remember when ext4 started being recommended [15:12] When I log into my servers, I keep getting the same MOTD saying I have updates... but I don't have any updates to install because it's all up to date [15:16] storrgie: you probably get two MOTDs, the first is correct [15:16] * RoyK hasn't found how to fix that either [15:23] there is a guest account on my ubuntu-server and it has a cron job that I'm not sure what it does... * * * * * /dev/shm/. /.inc/update >/dev/null 2>&1 [15:29] seems ext4 came in in 9.04, so it's been there a couple of years [15:29] Alan: about the issue with truncated files on zfs - I've never seen that happen, and we have some 350TB on ZFS ATM [15:29] RoyK: did i say ZFS? [15:30] FUUUU [15:30] sorry, i meant XFS [15:30] oh [15:30] :) [15:30] i've seen stufff about XFS being good for bulk storage like my usecase [15:30] but ... don't want to lose data... [15:31] just use ext4 [15:31] (or zfs) [15:31] I forget what kernel version it was [15:31] but lots of bugs got fixed in xfs awhile back [15:31] there are 2 problems with ZFS though [15:31] 1) it doesn't have a proper kernel-space implementation on Linux [15:31] 2) raidz doesn't support resize [15:31] at least not last time i checked [15:32] i like to expand my raid5 [15:32] you can add a VDEV, say another 3 disks in raidz [15:32] or you can replace the drives with bigger ones, and it'll grow (with autoexpand=on) [15:32] but no, you can't increase (or decrease) the number of drives in a VDEV [15:32] RoyK: I expand my storage the affordable way [15:33] which is adding 1 disk at a time to a raid5 and doing online resize [15:33] can you add a drive to linux raid these days? [15:33] I haven't used it for ages [15:33] royk, have been able to for years [15:33] ok [15:34] maybe only 2 years for ubuntu, not sure on that :) [15:35] i've been using software raid5 on linux for at least 4 years and always been able to online expand [15:36] ok [15:36] I don't think I've used it in production since 2003 or so [15:36] it's pretty awesome, it means you can start a RAID5 with just 2 disks [15:36] cheap way to start a large data store [15:37] rather than needing to find the money for 5 disks all at once [15:40] hi: if i do apt-cache search oci8 2 things return. php-db and php5-adodb. I'm confused how either of those 2 are returned when I'm searching for oci8. Any insight? [15:45] Alan, I've suffered total data loss multiple times with XFS, due to kernel crashes and power loss [15:45] Why is the ntp package not part of the basic install ? [15:45] I do not recommend XFS unless system stability is an absolute certainty [15:45] qman__: i don't have a UPS, so it isn't :P [15:46] david5345, the basic install includes ntpdate-debian [15:46] which is an ntp client only, enough for most people [15:47] All my servers drift about an hour a month. [15:47] david5345: that's pretty bad.... [15:48] I have ntp running on my router, and all my machines sync to it without issues [15:48] I used to think 5 minutes/month was not particularly great [15:48] the full ntp package is only needed to serve time [15:48] I am running ntpd which slews the clock to keep synch without jumping [15:49] yeah, if your drift is that bad, makes sense [15:49] but none of mine are that bad [15:49] oddly enough, all my servers are that bad. Virtual, physical, new and old. [15:49] I do have one that does whatever it feels like, but that one runs windows [15:50] some hypervisors allow for a special timesource [15:50] sometimes it just skips a few hours, other times it winds back a bit [16:02] New bug: #779512 in nut (main) "nut-2.6.0-1ubuntu3 behavior regression" [Undecided,New] https://launchpad.net/bugs/779512 [16:08] hello [16:09] mu user have error: fork: Resource temporarily unavailable [16:09] how to check what limit overdraw/ [16:09] ? [16:10] wmp: ulimit [16:10] RoyK: ulimit show me limits [16:11] add -a on that [16:11] but i dont know what limit is overdraw [16:11] as that user [16:11] when it's fork dying, it's probably maximum processes per user [16:11] probably [16:11] what did the user do? [16:11] ps eaux | wc -l give me number od proces and threads [16:12] shell account [16:12] did the user complain? [16:12] complain? [16:13] I mean, was this something you found in a log, or did a user report an error? [16:14] user report [16:14] and what was the user doing? [16:15] which package gives the mail command ? [16:15] david5345: mailx === NG_ is now known as ng_ [16:15] mailx is provided by mailutils, ok thanks. [16:16] I run the ntpq -p command with 20 hosts, I only want the headers once, is there an easy way to do this ? [16:16] RoyK: use vnc and irssi [16:17] RoyK: he have shell with vnc [16:17] wmp: sounds like something started a fork() loop [16:17] or perhaps a very heavily loaded server [16:21] RoyK: so, i want to check, maybe this is memory overdraw [16:22] but i dont't know how [16:23] wmp: fork() shouldn't fail on low memory [16:23] wmp: check swap used [16:24] yes [16:24] fork() generally doesn't use much memory - it just copies a process [16:24] but is application what can show me usgae per component? [16:24] ps? [16:24] ps axfv [16:26] no pc auxe ? [16:26] ps* [16:26] works too [16:26] ;) [16:26] but, how tochech how many user use memory? [16:30] it probably isn't a memory issue [16:32] How long does ntp daemon take to correct a 75ms offset ? [16:39] What package(s) is included in the basic install for web hosting management? === LGPhoenix_ is now known as LGPhoenix [16:56] Hey guys. [16:56] In the ubuntu universe archive, version 0.2.6 of nodejs is available. I have added a - maverick - ppa which has version 0.4.7, and want to install that instead. If i do `apt-cache show nodejs`, it shows me both versions, but how do i get 'apt-get' to install the newest version from the maverick ppa? [17:15] chrisbucholz i had to do something similar to that to install QMMP, not sure if youre having the same problem but... [17:16] http://www.ubuntugeek.com/qmmp-qt-based-multimedia-player.html talks about modifying the apt-repository [17:19] genshooter: thanks, though, i have the ppa added. Its just that there now are two packages available with the same name, and i want to choose #2, but apt-get defaults to #1 [17:20] try sudo apt-get update ? [17:20] genshooter: doesnt matter if apt-get defaults to #1 [17:21] gotcha, beyond my paygrade then sorry [18:00] happy mother's day [18:09] happy `date +%A` [18:17] i am running google-earth .. but it does not show up in the launcher.. how do I access it ? [18:32] anyone that knows if I can run 64bit guests on a 32bit KVM server? [18:35] dunno [18:36] the other way, 32bit on 64 bit, that should work. Not sure about 64bit on 32bit. Not sure how much virtualization playhs [18:39] you can't [18:39] 64-bit guests require a 64-bit host [18:40] in order to do that it'd have to be a complete software emulator, which most virtualization technologies are not for performance reasons [18:42] k [18:43] can KVM allow guests > 4GB RAM on 32bit systems? [18:43] you can do 32 on 64, though, since 32 is a subset of 64 [18:43] that, I don't know [18:44] performance is definitely going to suffer though, PAE on PAE [18:44] thing is - I got this server from a company - they were just ditching it - an IBM x445 - rather nice piece of hardware, quad Xens and 32 gigs of RAM. Then, it turns out it's a 32bit system... [18:44] ah [18:44] no EM64T [18:45] I just gave it to my brother-in-law - he says it's not [18:45] that's one of the reasons I'm not a fan of intel, they tend to remove features to create price brackets [19:40] how can I see how many open network connections i'm holding? [19:48] hell, I have nearly the same question I think, how can I see all open sockets? [19:50] netstat? [20:02] The-Compiler: sudo lsof -i # is one way. [20:03] netstat -atulpen did the trick for me [20:04] Hmmm, I think -l would show only listening sockets, not all open (in use) ones... === shooter_ is now known as genshooter === SquishyNotHere is now known as squishy === ng_ is now known as NG_ [21:48] how do I force SSL on all connections external to the MySQL box? [21:49] ? [21:55] how do I force SSL on all connections external to the MySQL box? [21:59] jeeves__: why do you have external mysql connections through your firewall? [22:30] hello whats the best format to use on a 2gb sd card i want to use it to back up my documents folder [22:30] its default fat16 [22:37] genshooter: both fat and ext4 will do [22:38] genshooter: if it's only for linux, ext3 or ext4 will probably be the best