=== dw3 is now known as dw1 === acrocity_ is now known as acrocity [03:14] Installing Ubuntu 13.10 Server to 890FX Motherboard with AMD Quad Core. Dying on Select Language Screen. [03:15] Any wise ideas before I begin google hunt for answers? [03:15] linux_clown: any chance you can try 14.04 LTS? it may have already fixed an issue [03:16] Well, I was troubleshooting a problem with Zentyal which uses 13.10 as base. [03:16] ah [03:16] Come to find it looks like 13.10 Server is the problem... [03:17] 'cuz now I load 13.10 server CD and same lockup... [03:17] Any of those F6 options worth messing with? [03:18] linux_clown: maybe try maxcpus=1 kernel command line option, maybe try adding 'nomodeset' kernel command line option [03:18] (both are blind guesses) [03:19] will try... can't hurt === dw3 is now known as dw1 [03:24] nope... still locking on the language window [03:25] linux_clown: darn. sorry. [03:25] nothing harmed. thx for the suggestion. [03:29] https://bugs.launchpad.net/ubuntu/+source/linux/+bug/1244176 [03:29] Launchpad bug 1244176 in linux "Server 13.10 Install Fails with USB Keyboard (Appears to Hang)" [High,Fix released] [03:29] USB keyboard - could that really be a problem? Dang... [03:30] Guess I need to hunt out a PS/2 keyboard in the attic... [03:30] linux_clown: wow, that looks like it exactly [03:31] Dang... at the office, too... can't believe there are no old PS/2 keyboards here... [03:31] yeah... even mentions failing at the select language screen.. [03:31] surely someone's got a model m or something? :) [03:32] got a friggin' keyboard museum in my garage and attic and storage unit... [03:32] hahaha [03:32] No worries... but don't want to wait to make the trip... [03:33] maybe there's some usb/ps2 adaptors laying around? [03:33] good idea... I know I have a bag full of those somewhere. Again, probably at home... [03:35] I guess not a problem with 14.04 server? [03:36] I hope not, but the next-to-last comment on that bug report mentioned he saw something similar [03:37] burning 14.04 right now - going to see... [03:46] god bless america... found a ps/2 keyboard in my buddy's desk drawer... [03:47] \o/ [03:59] that's absolutely the problem... keyboard... unbelievable... [04:00] crazy [04:03] booting like a champ now [04:09] linux_clown: Sometimes the BIOS has a USB legacy mode in which a USB keyboard pretends to be PS/2. Often there's a chicken and egg about actually selecting that option. (You need the keyboard to change it, etc.) [04:11] Typical BIOS error case in point: "You have a keyboard error, press F1 to continue." [04:14] hallyn,stgraber, solar designer wonders aloud if LXC rely upon SECURE_NOROOT: http://www.openwall.com/lists/oss-security/2014/05/01/3 -- the whole thread may be worth reading [04:17] maybe it's too late, but i'm not quite following [04:17] lxc isn't using SECURE_NOROOT [04:17] but if no_new_privs is badly interacting iwth it, then the problem isn't secure_noroot... [04:18] I haven't followed the thread closely enough to know if andy's on to something or not [04:18] andy lumo? he's in the thread? [04:19] * hallyn looks for a better view of the thread [04:20] yeah, luto's normally on the ball :) [04:27] i think i'll leave this ball o' fun for tomorrow, and finish up my unrelated patches tonight... [04:27] good plan === dw3 is now known as dw1 [05:26] "release-upgrade-motd" is consuming 100% CPU (http://askubuntu.com/questions/322343/check-new-release-process-eating-up-resources-on-ubuntu-server-13-04). I've calmed it down by changing /etc/update-manager/release-upgrades and setting Prompt=never, and changing /etc/cron.weekly/update-notifier-common commenting out #/usr/lib/ubuntu-release-upgrader/release-upgrade-motd. [05:26] Any idea as to why it's consuming 100% cpu when running? [06:09] Hi all. I'm fooling around with quota (for an sftp server I'm setting up). I have enabled journaled quota and a file aquota.user have been created on my /srv partition. My first question is if these 1k blocks thta df reports and that you can set for each user means 1024bytes? [06:11] jonascj: 'blocks' may be 512 bytes. it'd be worth checking to make sure you don't give out half of what you expect :) [06:12] So the quota database does not expect 1k blocks, just blocks. And the blocks can be whatever size? [06:12] empty folders on my system take up 4096 bytes, so isn't that one block? [06:15] jonascj: it'd either be 512, because that's the historical size, or it'd be 1024, because that's way more convenient :) [06:16] I really don't think it'd be 4k, even if that is the smallest 'sector size' the filesystem is willing to work with [06:16] sarnold: dumpe2fs /dev/vdb1 | grep -i 'block size' (my sftp data lives in /srv which is mounted on /dev/vdb1). That returns "Block size: 4096" [06:18] jonascj: hey! I finally found some reference: "For each user the current number of files and amount of space (in kilobytes) is printed" -- http://manpages.ubuntu.com/manpages/precise/en/man8/repquota.8.html [06:19] since the quota options have to be set in fstab for the individual partitions (which house file systems) I would think that the block size referred to by the quota system is file system block (whatever that means) opposed to harddisk sector size (maybe harddisks also have a block size, I don't know) [06:21] jonascj: hard drives do; drives smaller than 2tb are typically 512 bytes, larger drives and some newer drives are 4096 bytes, though some lie about it and pretend to be 512 bytes for the sake of ancient windows machines which can't fathom 4k drive sectors. I believe some ssds even have larger sizes still, not to mention their huge 1M or 4M 'erase blocks'. sigh. :) [06:21] but it cannot be harddisk block sizes that the quota system refer to, can it? I mean it is a filesystem feature, not a harddisk feature... [06:22] jonascj: it -could- be filesystem block sizes, but that'd be awkward to work with as an administrator -- you'd always have to check before using them [06:22] jonascj: yeah, exactly [06:23] jonascj: the only question is were they kind enough to give up on the traditional 512 bytes... [06:23] sarnold: so you think it is the harddisk block size / sector size? [06:24] jonascj: It's probably 1024 bytes. [06:24] or maybe it just is 1k blocks like the quota system says. All "tutorials" I've found speak of 1k blocks (1024bytes I assume). The quota system could work with that regardless of what the blocksize of the filesystem is. [06:25] my laptop's disk is currently being hammered by some long-running operations otherwise I'd go test it out :) hehe [06:25] wouldn't the easiest for administrators be to specify sizes in bytes? MB, KB, GB... [06:26] jonascj: yeah, but I wouldn't be surprised if these tools don't see much love :) [06:26] once they came up with the 64-bit size format they probably have'nt touched them since [06:26] :P [06:28] but I am still confused. Right now I believe the word "block" does not mean anything too special. It is just means 1024 bytes. So if I want a 10MB limit I would have to specify 10 [06:28] 10*1024*1024/1024 = 10240 blocks [06:28] right, that's what I would expect too [06:29] not 10*1024*1024/4096 = 2560 blocks even though dumpe2fs reports 4096 block size for my filesystem [06:30] I guess I'll just test it. Although I do not know how fast one should expect the quotas to be updated / inforced. [06:30] I -hope- they'd be enforced near instantly [06:30] or if I need a manual cronjob to update the quotas or if it is done automatically upon filesystem activity (the last option qould be the smartest) [07:01] sarnold: it turned out to be blocks of 1024 bytes. Setting 4096 as hard limit allows me to upload 4MB of data. repquota is 10sec delayed or something === negronjl-afk is now known as negronjl [07:22] Good morning. [07:31] if the owner of a file do not have read permission but his group or other have read permission, do the user have read permission after all? [07:31] jonascj: If the user is in a group with read permission then the user can read the file. [07:32] and if he were not in such a group but other (owner-group-other) had read permission, the user would still have read permission? [07:32] yes. [07:33] If other has read access, everyone on the system can read the file/folder. [07:37] good, then I cannot prevent the owner from doing anything but allowing others to do said thing using basic permissions :) [07:37] * or not good, it would be useful for me right now, but it wouldn't make sense if it worked that way. So good after all === Lcawte|Away is now known as Lcawte === deegee__ is now known as drussell [08:26] hey all...would someone maybe know if LTS 14.04 ISO's already contain "out-of-the-box" patched version of openssl or do I have to remove/update it after clean installation? [08:27] you have to run updates. [08:27] Allways a good idea to run updates ;) [08:29] yeah...from my previous experience running updates on production server wasn't "always" a good idea... but anyway [08:30] ah of course...does openssl comes preinstalled with default instllation or can you install it later fresh/updated version? [08:31] !info openssl [08:31] openssl (source: openssl): Secure Sockets Layer toolkit - cryptographic utility. In component main, is standard. Version 1.0.1f-1ubuntu2 (trusty), package size 468 kB, installed size 899 kB [08:31] ^ makes me think it is preinstalled. [08:31] seems so [08:32] thanks for answers! [08:33] https://launchpad.net/ubuntu/+source/openssl/1.0.1f-1ubuntu2 [08:34] that's the link I found...and here it says it's already patched against heartbleed [08:34] is that one in the ISO's or not? [08:34] keglevich: ^ is in the repo, I'm not sure what version is on the iso. [08:34] aha ok, I'll check it out then [08:35] or, is there any way to check it out before installing..I mean the "repo" version? [08:36] http://askubuntu.com/questions/449184/how-to-upgrade-openssl-1-0-1f-on-ubuntu-server-14-04 [08:36] here again someone says it already contains updated version [08:36] 1.0.1f-1ubuntu2 is on the Ubu desktop iso. [08:37] so probably server ISO's contain same version [08:38] I guess.. [08:46] what happens to an expired account (e.g. an account created with "-e yyyy-mm-dd")? Will it remain in place with all it's files and everything but just with an invalid password? [08:57] $ rmadison -strusty openssl openssl | 1.0.1f-1ubuntu2 | trusty | source, amd64, arm64, armhf, i386, powerpc, ppc64el [08:57] keglevich: ^^ this tells me that it's on the ISOs. If it weren't, I'd see trusty-updates, and it'd be on the next point release. [08:58] For 12.04, it looks like the next point release will be in August. [08:58] jonascj: correct [09:02] http://www.reddit.com/r/cablefail/comments/24cxa7/50meter_lan_no_idea_if_this_works/ === Lcawte is now known as Lcawte|Away [09:24] rbasak: thanks [09:28] the passwd is ok on an expired account, you just can't log in with it :D once you unexpire you can use the same passwd to log in (just to be clear) [10:42] hi [10:54] I would need a hand with my private/local debs repo [10:54] does anyone has experience with build such? [10:57] !anyone | luk__ [10:58] bot or something? [10:59] Apparently that snippet was delted from the bot. [10:59] Anyway, the point is: just ask. [10:59] apparently ... well nothing is perfect [10:59] sure ;) [11:00] I have local repo defined as: deb [trusted=yes] file:/opt/debs precise main [11:00] today I've created new package (3rd one) and added to it (just cp and I've regenerated Packages.gz) [11:01] and when I'm trying to install this new package I get .... 1 sec ... [11:02] hmm ... now works :) [11:02] I think my package name was incorrect [11:02] :) [11:02] or to be more precise ... the version: 1.1-SNAPSHOT [11:03] probably '-' in the version is not the best idea [11:04] hmm .. strange, because i.e. my other packages has version like that: 1.5.1-1ubuntu1.1 [11:05] '-' divides the upstream revision from the packaging revision. [11:05] It has special meaning, distinguishing native and upstream packages. [11:09] hmm ... now SNAPSHOT work ... maybe the reason was that before I've added version 1.1-SNAPSHOT to repo I've first installed/removed it by dpkg -i/-r [11:10] Hi, are there any online courses for ubuntu server admin? [11:10] rbasak: what is the meaning of native and upstream? [11:11] Googles it all the links just seem to go back to the ubuntu support page === jrgifford__ is now known as jrgifford [11:20] rbasak: it was my typo I think, thanks for help anyway [11:20] luk__: it's where the source comes from [11:20] upstream is external [11:21] native is part of the same source tree, for packages that don't have an upstream (eg. dpkg) [11:53] morning [13:17] jamespage: we should be able to drop the python 3.2 compat patch in python-pbr right? === mjohnson151 is now known as mjohnson15 [15:31] shazbut [15:32] jamespage: awesome...just awesome.. [15:35] zul, wassup? [15:35] jamespage: utopic sqlachemy is 0.9, migrate still depends on 0.8 [15:39] any one any experience in setting up transmission on ubuntu server === Lcawte|Away is now known as Lcawte [16:27] zul, that's making all the openstack packages un-installable right? [16:32] jamespage: working on it :) [16:32] need some new deps MIR as well [17:25] hi - looking for some simple advise about optimising LVM to house QCOW images. [17:25] I am creating a KVM server that has a 3.2 TB VG to house QCOW images - i.e /var/lib/libvirtd [17:25] i.e a 3.2 TB LVM volume group - I created the volume group with the Ubuntu-server installer and it defaulted to 4MB PE Size [17:26] would a VG of this size (and for this task) be o.k with 4MB or would increasing the size gain better performance? [17:28] 4M is usually okay, since the block size of qcow images is smaller than 4M. [17:29] bekks: thanks for the clarification [17:29] until now I have always housed KVM vm's in LVM volumes (instead of disk images) [17:30] but now we can do live snapshots with images.... [17:52] Hello [17:52] Please view this : http://paste.ubuntu.com/7374561/ [17:52] why system startup check for stoping some service on startup ? [17:53] for example : Starting load fallback graphics devices [17:53] and then : Stopping load fallback graphics devices [18:27] hi guys, anyone else here using nfs-rdma? On 14.04 it seems to be extremely kernel-panicky [18:31] it worked fine for me on 10.04 [18:31] I haven't used it since then === negronjl is now known as negronjl-afk [19:31] Hey guys, [19:31] I'm getting [19:31] GPG error: http://us-east-1.ec2.archive.ubuntu.com precise-updates Release: The following signatures were invalid: BADSIG 40976EAF437D05B5 Ubuntu Archive Automatic Signing Key [19:31] https://forums.aws.amazon.com/thread.jspa?messageID=530139 [19:31] that's an awful long DNS propogation time :) [19:31] *propagation [19:44] anyone run into ubuntu servere renaming ethernet interfaces and ignoring 70-persistant-net-rules [19:44] I think it has to do with this worthless pile of crap known as networkmanager [19:45] iscsinoob: n-m on a server? [19:45] isn't not there? [19:45] oh then nm [19:45] but it's still ignoring 70-persistant-net-rules [19:46] lemme try something first [19:52] iscsinoob: it's probably because of biosdevname [19:52] ugh [19:52] so what's the solution [19:52] I am doing mac address matching in the file [19:53] I think you can boot with the biosdevname=0 kernel boot parameter if you want [19:59] i'm fine with the bios n ames but when I tried to setup networking it renamed one to rename4 [19:59] then I had to edit the config again [20:01] but lemme try renaming them [20:03] iscsinoob: honestly, I'd probably just kill your 70-persistant-net-rules file, and reboot [20:03] iscsinoob: and use whatever names come up as default [20:03] iscsinoob: with biosdevname, they shouldn't change anymore as they'Re based on slot now [20:05] that's what I would think [20:05] but as soon as I added stuff tin /etc/networking/interface it renamed it again [20:05] it's not consistant [20:07] I disabled biosdevname and lets see what happens [20:11] yeah it worked [20:11] soon as I did the biosdevname it renamed an interface to "yomamma" [20:15] what the heck was the purpose of biosdevname anyways [20:29] looks good, thx [20:52] iscsinoob: the idea is to make it easier to determine which NIC or hard drive is which in a consistent fashion when you've got thousands of machines racked up [20:52] ah [20:52] it feels like the kind of thing that would have been nice to have done a decade ago :) [20:52] Hi What is the best way to use a group with apache for manage website? [20:55] b.v. i have a group foo that have only acces to /var/www/foo and not acces to /var/www/bar/ [20:57] www2: you can definitely do that; the "BSD groups" behaviour, where you set the setgid bit on the directory /var/www/foo, might be useful too [21:01] sarnold: thx and where can i find more infomation about this? [21:03] www2: hrm, the best documentation I can find on short notice is in the mount(8) manpage -- thogh you don't need to change any mount options to use it [21:16] sarnold: oke i dig soon the mound page btw is it oke to add group foo to www-data user? [21:17] www2: yes, that should be fine [21:20] sarnold: oke i do it soon when i create the group [21:37] hallyn: I've got a pile of failed VMs for 1292234 -- would it be helpful to you for me to hold onto them all? or can I reclaim the space? [21:38] bug 1292234 [21:38] Launchpad bug 1292234 in qemu "qcow2 image corruption in trusty (qemu 1.7 and 2.0 candidate)" [High,Confirmed] https://launchpad.net/bugs/1292234 [21:38] btw can anyone tell me why pad.lv isn't working? [21:39] sarnold: hm, you've got a pile? do you have a regular recipe for reproducing them? [21:40] hallyn: not 'regular', really, but I started with iirc ten vms and now only two or three of them boot ;) [21:41] sarnold: well fooi. i'm not sure i'll be able to easily d/l them from hotels over the next bit... but anyway if you can just keep one around, maybe put it on p.c.c ? [21:41] hallyn: yikes yeah hotels aren't made for four gig transfers are they? :) [21:41] or if you know someone going to ODS or to austin next week, hand them a usb stick? :) [21:41] hhaha :) [21:51] sarnold: by any chance are you seeing gaughen or someone at a beer outing on sat? (you're there right?) [21:51] sarnold, there is beer being drank tomorrow but it's in beaverton [21:52] sarnold, I know how it's hard to venture that far west [21:52] it -is- hard to venture that far away from the couch :) [21:52] even for beer? [21:52] hallyn, I'd be happy to meet sarnold for beer on Saturday but he'll have to meet me at a children's birthday party [21:53] lol [21:53] sarnold, there will be cake [21:53] bouncy castle? :) [21:53] sarnold, nah, bouncy castles are so 2005, now it's arts and crafting [21:53] with reclaimed materials [21:53] oh man I won't get the glitter off my clothes for weeks! [21:53] hallyn: actually it bzip2's down to just 1.7 gigs, not nearly as bad as I feared [21:53] is that a bad thing? [21:53] lol [21:54] uh, i'm still all about the bouncy castles, at any rate [21:54] hallyn, you are so 2005 [21:55] sarnold: well, actually i should be able to fetch it from some cloud instance, so should be fine. if you can just post one to p.c.c i'll take a look [21:55] else, i'll try out your recipe next week [21:55] hello, i was wondering if the linux-virtual kernel is no longer available in 14.04? I can't find much info about this. [21:56] I'm trying to preseed VMs about continue to get the generic kernel [21:56] *but continue [21:59] repost from #ubuntu: Question.. I've looked online about this issue but none of the proposed fixes did any good. On my ubuntu 12.04 server, mysql fails to start on boot. I attempted to reset the init scripts with update-rc.d, I tried removing/adding the bind address, setting the on-start to wait for eth0, but nothing is working. Any ideas? [21:59] Second issue, at random, I get this strange issue of ksoftirqdx eating 100% of all cores on the system despite low usage/load on any of the services on the machine. [21:59] I've got "d-i base-installer/kernel/override-image string linux-virtual" as well [22:00] vonsyd0w: yeah, apt-cache show suggests that linux-virtual now just depends on linux-image-generic [22:00] Term1nal: does it start manually? [22:01] vonsyd0w: it's a transitional package in newer releases -- I assume it isn't needed any longer: http://packages.ubuntu.com/trusty/linux-virtual [22:01] heya guys is there a metapackage for a openstack testing environment? [22:01] wanting to doa one box wonder for play [22:02] pmatulis: yes, I can do: service mysql start [22:02] and it works just fine [22:02] thebwt: investigate juju openstack charms / bundles [22:02] hallyn & sarnold, ah OK thanks for the info. Makes sense. Thanks. [22:04] rgr, thanks sarnold [22:04] Term1nal: i guess you need to debug the init script. not sure if that means upstart or sysvinit [22:05] Term1nal: do you have any non-standard init/upstart scripts, change anything, etc? [22:05] thebwt, there is also devstack @ http://devstack.org - just so you have options [22:05] http://pastebin.com/u8LLsfV0 [22:06] pmatulis: not that I know of, just a pretty bland install of mysql on 12.04, I haven't muddled with any startup scripts until this issue started to occur. [22:06] thebwt: also, check out RDO and Packstack [22:09] thanks! [22:16] pmatulis: does it help that I'm also running postfix? mysql fails to start, then postfix won't work properly until I manually start the mysql service, then restart postfix. [22:17] hallyn: people.canonical.com:~sarnold/sec-precise-amd64.qcow2.bz2{,.sig} :) [22:19] sarnold: thx [22:19] hallyn: thank you! :) === Ursinha is now known as Ursinha-afk === Ursinha-afk is now known as Ursinha [22:53] Term1nal: interesting. is postfix using mysql as a backend? [22:54] pmatulis: correct. [22:56] pmatulis: Not sure why it's not working on start. Nothing in the syslog indicating -why- mysql isn't starting. dmesg just has generic exit with status 1 messages attempting to start it, and mysql error.log only has generic info messages and nothing critical as far as I can tell. [22:57] Term1nal: did mysql install without any issue at all? [22:58] it's been running for like a year no issues. [22:58] I just rarely restart the machine [22:58] it still works no problem functionally, it's just the startup on boot that is the problem. [22:59] Term1nal: so you don't remember whether there was any issue (at all) when *installing* mysql? [22:59] we've got a heat wave here in cali and our server room AC burned out so a few of our boxes thermal shutdown on us [22:59] one of them happened to be this particular ubuntu box :D [22:59] pmatulis: None whatsoever :( [23:00] well there's an update for mysql-server so maybe I'll give that a try :D [23:03] Term1nal: you'll need to investigate how to debug the mysql upstart job [23:04] Term1nal: http://goo.gl/TYm7ch [23:05] well whatever I tinkered with... [23:05] I just ran a dist-upgrade and rebooted [23:05] started up fine O.o [23:05] Term1nal: good for you [23:06] lol [23:12] If I have multiple dns servers in /etc/resolv.conf, will they be roundrobined? Or just failed over when needed? [23:12] failover [23:18] sup === mikal is now known as stillbert [23:20] heh === RaptorJesus_ is now known as RaptorJesus === mirelle is now known as cryptosolution === cryptosolution is now known as mirelle === Lcawte is now known as Lcawte|Away