[00:18] Yea, I filed the bacula bug. apparently, only x86_64 is correct. ldd fails to find libmysqlclient.so on i386, armel, and armhf, possibly ppc as well. [00:40] GrueMaster: should be looking for libmysqlclient.so.18 [00:40] Yes, it should. [00:41] the way bacula supports multiple db backends is kind of silly I think [00:41] (see the bug report for the gory details). [00:41] they should have figured out dlopen() years ago [00:45] It looks like we made a half hearted attempt to fix multilib issues in 5.2.5-0ubuntu1 by adding the path for mysql lib64. I say half-hearted as the patch failed to add support for any other arch. [00:45] GrueMaster: *DOH* === sixstringsg|away is now known as sixstringsg === sixstringsg is now known as sixstringsg|away === pdtpatr1ck_ is now known as pdtpatr1ck [02:56] New bug: #949608 in samba (main) "package winbind 2:3.5.8~dfsg-1ubuntu2.3 failed to install/upgrade: subprocess installed post-installation script returned error exit status 127" [Undecided,New] https://launchpad.net/bugs/949608 [03:32] hi guys, I'm trying to install ubuntu on my first raid setup. raid1 [03:32] it gets to partition area and asks for iscsi stuff? [03:32] nOStahl: md or fakeraid [03:32] I setup the raid in bios [03:32] not sure what md stands for [03:33] k I'm at screen now says ! detect disks / one or more drives container serial at a raid configurations have been found, do you wish to activate these Raid devices? === skrewler_ is now known as skrewler [03:34] nOStahl: do not set up raid in the bios [03:34] why not [03:34] nOStahl: because it is fakeraid [03:37] ok so go back into bios and blow away the raid stuff [03:37] Yes [03:37] then restart the installer? [03:37] set it back to AHCI [03:38] always set it to AHCI not hybrid or traditional or raid [03:38] learning something new everyday :) [03:38] it has option for SATA EMULATION legacy mode IDE or Native Mode IDE [03:38] in my bios [03:38] or RAID [03:39] its an hp dc5750 microtower [03:39] How old is this box? Several years? [03:39] 2006 [03:39] i'd say [03:39] probably predates AHCI [03:40] Try native mode [03:40] noticed the bios for it is up to 2.36 and its running 2.25 [03:42] I am usually pretty wary of firmware upgrades [03:42] I tend to avoid upgrades unless there's a problem you KNOW it will fix [03:42] aye I don't like chancing a brick [03:42] s/know/are reasonably confident/ [03:43] almost back to partition area [03:46] so it will let me setup the raid1 here soon? [03:47] k it is asking me again if I want to activate serial at a raid devices [03:48] you do not [03:48] though you may have to zero the disks to fix this problem, I ran into it [03:48] even though I said no, it detected and assembled anyway [03:48] and screwed everything up [03:48] k [03:49] so now at partition disks screen [03:49] guided - use entire disk or guided - use entire disk and setup lvm etc [03:49] or manual [03:49] if you want raid you want manual [03:49] k [03:49] so at screen now showing the two 1.5 terabyte hard drives [03:50] to do md raid, you have to create partitions of type 'volume for RAID' on each disk, then create the raid, then create partitions on the raid [03:50] you may want to create a /boot or swap area outside of the raid on the bare disks, though [03:51] hrmm [03:51] i am setting this tower up as a cloud controller / storage controller [03:52] with two dell optiplex 760 towers as nodes [03:52] so this box will be serving the images for the instances etc. [03:59] ok so I need to setup how large of a partition for /boot [04:04] qman___: the wiki does not say to select volume for RAID it says to partition each drive as normal and then go back and select configure software raid [04:04] does that sound right? [04:05] qman___: yes, easiest solution is to write zeros to the first and last ~16MB of disk === sixstringsg|away is now known as sixstringsg === jtv1 is now known as jtv [04:43] da/win 6 [04:53] roaksoax: ENOCHAN [04:59] lol [05:15] my ubuntu squid server with two NIC doest not work [05:15] hi all [05:16] my squid ubuntu server with two NIC doest work. one is for wan IP and one is LAN [05:16] my squid ubuntu server with two NIC does not work. one is for wan IP and one is LAN === philipballew is now known as hello_my_name_is === hello_my_name_is is now known as philipballew === sixstringsg is now known as sixstringsg|away [07:45] hello. i need that a couple of packaged be of newr version [07:45] what could i do? add mirrors for later versions of ubuntu? i am on LTS [07:58] also i found out there is some update-manager? [08:01] is there hope i can update the server without formatting? [08:11] Anaphaxeton: do-release-upgrade [08:12] No new release found [08:12] Anaphaxeton: that'll upgrade the release to whatever's next, or to the next LTS if running an LTS, that is, if running 10.04, it won't do much, since 12.04 isn't released [08:12] :( [08:12] ok [08:12] Anaphaxeton: running lucid? [08:12] is lucid 10.04? [08:12] if so, you can choose to upgrade to a 'normal' release (default is next LTS release) [08:12] yes [08:13] i used the -d flag [08:13] if you want to upgrade to 10.10, change /etc/update-manager/release-upgrades and set prompt=normal [08:13] erm... [08:13] that'll upgrade to 12.04 *beta* [08:13] which may not be a good idea [08:13] yes [08:13] maybe [08:14] but it's up to you ;) [08:14] however i am not using it for exotic stuff [08:14] then it'll probably work [08:14] a simple home server for irc logging and torrents [08:14] keep a backup in case the midden hits the windmill... [09:08] hi. how can i check if I have packages installed from outside the standard repos main/universe/multiverse? like PPAs, third party repos etc? [09:09] apt-cache policy foo, might help [09:13] New bug: #946953 in samba (main) "'Permission denied' when using a windows share" [Undecided,New] https://launchpad.net/bugs/946953 [09:18] twb: I would have to do this on every single package I have installed. isn't there a more efficient way? [09:22] morning o/ [09:26] cemc: write a sh script? [09:44] i ve done it [09:45] i destroyed my server [09:45] the upgrade process could not complete [09:45] and a reboot made the system useless [09:46] then running a usb stick distro [09:46] seems to not see mt raid 0 [09:46] my* [09:49] hello all [09:49] hey guys, im sorry for the n00b question, but can someone explain to me what does ubuntu cloud means exactly? [09:50] My Ubuntu 12.04 A2 and B1 installs keep kernel panicking on boot: log.c:768: Assertion failed in log_clear_unflushed: log->remote_closed (Kernel 3.2.0-17-generic) [09:50] drag0nz, http://en.wikipedia.org/wiki/Cloud_computing [09:51] anybody else see this - was hoping it was an Alpha-2 issue. [10:14] drag0nz: it's a buzzword. It basically means the same thing as time-sharing (60s) or thin clients (90s) [10:15] drag0nz: its ubuntu's ability to provision cloud IaaS services out of the box [10:19] hey folks [10:20] I'm trying to script the creation of LXC containers, using disk image files [10:20] is there a way to get mkfs.ext4 to not ask if the device specified is not a block device? [10:21] s/ask/ask for confirmation/ [10:21] what is the recommended way to restrict ssh access to only scp/sftp (no shell) on lucid 10.04 ? [10:21] I tried -q, but that doesn't seem to make a difference [10:21] cemc: http://www.debian-administration.org/articles/94 comes to mind [10:22] kai: thanks! [10:23] first hit for "ssh sftp only" on google, btw :) [10:24] cemc: http://solderintheveins.co.uk/2011/03/ubuntu-sftp-only-account-how-to/ has a bit more comprehensive setup [10:26] thanks [10:26] no worries :) [10:29] cemc, you may have to add a Match to close the Match Group "directive" in sshd_config [10:37] hi all [10:37] bluefrog: thanks for that. the manual says it's either another Match to close the block or the end of the file. since the guide adds those at the end probably that's why it doesn't need a closing Match [10:51] lynxman: do you have access to any amd servers? [10:52] Daviey: yessir and they're even running Ubuntu [10:53] lynxman: \o/, precise? [10:53] Daviey: Oneiric, they're prod machines [10:53] Daviey: AMD Opteron(tm) Processor 242 [10:53] Daviey: was thinking about upgrading them to precise though [10:54] lynxman: dammit, any lab machines running precise on amd64, that you know of? [10:54] * RoyK has one or two [10:54] Daviey: all of them are xeon [10:55] pentium III xeon? ;) [10:55] Daviey: as said, I can upgrade them to precise, its my mysql prod db [10:55] Daviey: nothing wrong can happen right? *wink* [10:56] lynxman: really, upgrading a production machine to precise? now? [10:56] RoyK: I'm reckless like that [10:56] RoyK: it's actually an active-passive cluster, so if I upgrade just one should be fine [10:57] and you're not afraid that the older version of mysql will get a nervous breakdown when its peer has been upgraded and data on disk changed to a new format? ;) [10:59] eeek [10:59] RoyK: Yeah, i really want to get some data from an amd64 machine on precise :( [10:59] thanks anyway [11:00] RoyK: nah ;) [11:08] upgrading an ubuntu server is a pita... [11:08] goodbye... going to my desktop distro, archlinux! [11:08] oh divas :) *shrug* === mrevell_ is now known as mrevell [11:42] Daviey: I think I'm misunderstanding the triage report - http://reports.qa.ubuntu.com/reports/ubuntu-server/triage-report.html [11:42] when I set the importance I would expect bugs to move from New/Undecided to New? [11:42] is that not correct? [11:50] jamespage: it should... [11:52] Daviey: is there a nice way to make apt-get be more verbose on about what happens in maintainer scripts [11:52] error code 11 is not very descriptive... [11:52] jamespage: set -x ? [11:53] Daviey: doh! [11:53] jamespage: dpkg --debug=1000 [11:53] * jamespage faceplants [12:32] jamespage: if I could get a patch that fixes bug 949044 and touches only ARM specific code, this could get sponsored through feature freeze, right? [12:32] Launchpad bug 949044 in openmpi "Basic openmpi hello world fails on arm" [High,Triaged] https://launchpad.net/bugs/949044 [12:33] rbasak: how is that a feature? [12:33] Daviey: that's what I mean. It's a bug, so should have no issues getting uploaded? That's my understanding, I just wanted to check! [12:34] rbasak: right! :) [12:51] hey zul [13:14] apart from changing the middleware configs at api-paste.ini does packages(services) nova/api/volume/glance really needs some more changes after the new keystone? === AndroUser2 is now known as emkoemko [13:20] hi [13:51] rbasak, as Daviey says - that should not be an issue [13:51] jamespage: great, thanks [14:05] what is the url that one shoudl use to be able to have a server doa network install from the archive? [14:06] is this the url that i want http://archive.ubuntu.com/ubuntu/dists/lucid/ [14:12] lynxman, are you doing some updates for puppet? [14:13] Daviey: i havent been able to reproduce that lxc network injection bug [14:16] hey guys, I'm setting up my first raid a raid1 setup with two 1.5 terabyte hard drives, should I also do LVM? [14:18] zul: :/ [14:19] * smb guesses this is not a particularly good time to nag zul about reviews/sponsoring... [14:20] smb: xen right? [14:20] zul, Right, though I have another one for libvirt by now... [14:20] bug 948333 and bug 949028 [14:20] smb: queue them up right now before i forget :) [14:20] Launchpad bug 948333 in xen "Xen: pxeboot for e1000 emulation not working" [Medium,In progress] https://launchpad.net/bugs/948333 [14:20] Launchpad bug 949028 in ubuntu "libvirt: xen: never use type=ioemu for vif definitions" [Medium,Triaged] https://launchpad.net/bugs/949028 [14:23] smb: doing it right now [14:23] zul, many cheers :) [14:25] <_ruben> eagles0513875_: http://archive.ubuntu.com/ubuntu/ oughta be enough for a netinstall url [14:25] _ruben: well xenserver is complaining when it tries to update the archives saying there is no release file [14:26] that is just a http reference to the URL [14:26] you need more info that that [14:27] ikonia: like what version etc [14:27] jamespage: fancy reviewing a srcNEW? [14:28] Daviey: OK [14:28] correct [14:28] ok thanks ikonia [14:29] jamespage: http://people.canonical.com/~davewalker/convoy/ [14:29] * jamespage goes to grab coffee [14:30] whats convoy? [14:31] Hello - i am needing to setup an smtp server - before i get to far in - am i on the right path with using Postfix for this [14:33] zul: https://launchpad.net/convoy [14:33] New bug: #949956 in lxc (universe) "lxc-start-ephemeral keeps same MAC address, and so same IP address, as base instance" [Undecided,New] https://launchpad.net/bugs/949956 [14:33] Daviey: oooooh... [14:34] smb: did the libvirt patch come from upstream or totally original [14:35] zul, Done by me [14:35] smb: ok just wondering [14:35] zul, It might be something to feed back [14:36] smb: i was just thinking that [14:36] zul, Just not sure what the "usual" procedure would be [14:37] smb: i would say talk to hallyn since he maintains it but he isnt here this week [14:37] smb: anyways uploaded both [14:37] zul, Ok, will do so. Great, thanks [14:40] hey guys, i installed ubuntu server and now when I reboot it show grub and then its blank right after that.. [14:40] ubuntu 11.10 [14:41] zul, Btw, the xen patch also might be something to feed back to Debian. I think the change of rom file was part of a Debian patch. [14:42] smb: ack [14:42] any ideas? [14:45] jamespage: yes, adding some patches that are required to revert a previous behaviour in lockfile treatment [14:45] jamespage: almost done, it's a bit of a pita :) [14:45] lo. [14:45] lynxman, can you pickup a couple of fixes for bug 948983 [14:45] Launchpad bug 948983 in puppet "puppetmaster-passenger default vhost has wrong documentroot" [High,Confirmed] https://launchpad.net/bugs/948983 [14:46] I think they are both trivial - glenn_ is the reporter [14:46] jamespage: glad to do so :) [14:46] nOStahl, Could be various things. You could try to user "nomodeset" on the kernel arguments from grub. I think 11.10 did not use quiet anymore... [14:46] jamespage: as i was saying, it might be a good idea, that the puppetmaster-passenger checks if there is a /etc/puppet/puppet.conf before installing, and either backups that file or asks the user for confirmation [14:46] k so I'm in grub and I hit e [14:47] glenn_: I reckon that's done automatically by the packaging for any config file, right jamespage? [14:47] shows things like setparams and ins mod stuff [14:47] where do I do nomodeset [14:48] glenn_, ah - so that file is provided by puppet-common [14:48] nOStahl, Should be the line which also sets the root= I believe, give me a sec [14:48] ty smb [14:48] and lynxman is correct in that any changes to that file made locally are preserved by default be the package management system [14:48] you can elect to take the new file - but its an interactive prompt [14:49] jamespage: i didnt got that question, but perhaps thats because of the error in the package? [14:49] nOStahl, Actually the one starting with linux and with root=xxx ro <- there [14:49] hmm [14:49] if I've created my own package repo, how do I sign it with gpg? [14:50] it should be classified as a conffile [14:50] so it should prompt [14:50] in the < >? [14:50] jamespage: will tackle both at the same time and repush the branch, if you fancy reviewing later I'll be honoured :) [14:50] nOStahl, no, sorry, without [14:50] k ty [14:50] lynxman i can test if you want [14:50] lynxman, ack - happy todo so [14:50] glenn_: oh that would be superb :) [14:50] jamespage: yay \o/ [14:50] lynxman if youd put it on ur ppa i can test it [14:51] lynxman, I need to look at that rabbitmq package again don't I [14:51] did that for the backports of lucid/./. too [14:51] glenn_: will do so [14:51] glenn_: thank you very very much! [14:51] * jamespage pokes himself on behalf of lynxman [14:51] jamespage: yeah :) [14:51] smb: k did that then hit f10 to boot and still black screen [14:52] nOStahl, f10?... not sure I user ctrl-x to boot a changed setup. but maybe I just ignore other keybindings [14:52] jamespage: so, if id have /etc/puppet/puppet.conf created on a fresh server, you say that if the package install wants to replace that file it should prompt the user? [14:52] ah ok [14:52] restarting lets check [14:53] k tried it again hit e and added nomodeset to the end of ro line [14:54] ctr-x and still black screen [14:54] nOStahl, Yes, actually says both do the same [14:54] this is on a fresh install of ubuntu server 11.10 raid1 [14:54] nOStahl, Just had never relaized it changed or was added [14:56] what do I do now heh [14:56] nOStahl, you could add "debug" instead of nomodeset, which will make more boot messages visible [14:57] just tried recover mode kernel [14:57] nOStahl, And what kind of raid1 (hw, fakeraid/dmraid or softraid/md)? [14:57] its booting [14:57] softraid/md [14:57] dropped me saying degraded drive [14:58] softraid.. UGH [14:58] SpamapS, purist. :) [14:58] nOStahl: you should be able to say "Y" to boot degraded [14:58] smb: apologist! [14:58] SpamapS, Never! :-P [15:00] its saying things like md0 unknown partition stuff [15:00] Anyway, that actually reminds me that I am not sure how smart grub nowadays is, but I'd always have /boot on a non-raided partition. I think normally the installer does that automatically [15:00] mdadm CREATE user root not found [15:01] nOStahl, unknown partition may be if the whole md is used a lvm pv [15:01] I didn't setup lvm [15:02] jamespage, lynxman: the install of the puppetmaster-passenger package asks me if it should replace the config file [15:03] ya its just looping trying to start in degraded mode [15:03] glenn_, puppetmaster-passenger should not touch /etc/puppet/puppet.conf [15:03] but it will pull in puppet-common which will want to update that file [15:03] reinstall is in order it looks like [15:03] puppet-common that is ofcourse, as dependency [15:03] update/replace [15:04] who is squigley.namespace.at ? [15:04] i got that in my puppetmaster vhost now :) [15:04] nOStahl, Hard to say... trouble is that it is rather rare I have to resort to mdadm command. so I forget [15:05] nOStahl, I think there was a command to inspect the individual disks meta data... [15:05] mdadm --excamine /dev/xxxx [15:05] examine even [15:05] SpamapS, Right, thanks [15:05] nOStahl: what happened? RAID1 problems post install? [15:05] ya [15:06] * SpamapS tests that every release, and so would be interested to hear if there are holes in the testing [15:06] I followed the ubuntu server setup guide [15:06] I'm running an hp dc5750 with two caviar black 1.5 terabyte hard drives [15:06] lynxman: can you double check if that hostname is in the puppetmaster-passenger code? it seems hardcoded [15:07] lynxman: on my server i have the hostname of a developers machine [15:07] nOStahl: the degraded boot should drop you to a shell if you say N [15:07] SpamapS: should i do that first before trying a reinstall? [15:08] I think that would be helpful to find out what went wrong [15:09] k one min [15:10] nOStahl: yes, there may be a bug that needs fixing.. it would help a lot. :) [15:10] nOStahl: is this 11.10 ? [15:10] Daviey: convoy feedback - http://pad.ubuntu.com/convoy [15:10] k at initramfs [15:10] 11.10 yes [15:11] nOStahl: so first I'd just do mdadm --examine on each device you think should be part of the array [15:12] example? [15:12] mdadm --examine /dev/sda [15:12] if you have an sda [15:13] k ty [15:13] hey guys, if install the tomcat java server do i need to install the java JRE? [15:13] no md superblock detected on /dev/sda [15:13] same with /dev/sdb [15:13] df shows no devices exempt for udev and tmpfs [15:14] nOStahl: partitions not devices [15:14] nOStahl, And when you created the raid, you used sda and sdb or sda1/sdb1 [15:16] k walk me through this lol df shows no devices except for udev and tmpfs [15:16] http://paste.ubuntu.com/874680/ [15:16] so mdadm can't find any devices to check? [15:16] nOStahl, what does cat /proc/partitions show? [15:16] checking [15:16] sda / sdb / dm-0 / dm-1 [15:17] nOStahl, hm the dm-* ones are surprising... dmsetup ls --tree (hoping this works) [15:18] that spit out pdc_bchdbda1 [15:18] and some other lines [15:18] that reeks a bit os fakeroot.... [15:19] nOStahl, could you paste me the whole output as a pm or pastbin [15:21] done [15:22] So that looks like a bit like a dmraid consisting of sda and sdb and probably a partition on that... [15:23] If you could paste dmsetup table [15:25] So yes the pdc_bchdbda1 is a partition on a mirror. But not managed by md [15:25] nOStahl, Can it be that you also have some RAID1 option set up in the BIOS [15:26] I did at first [15:26] but then i put it back to normal [15:26] as I was told I didn't want the bios version [15:27] nOStahl, Ah, that is potentially the problem. Maybe meta data on the devices was not removed properly. So now dmraid still grabs the devices [15:28] boot up a gparted disk and zero them out or something and reinstall ? [15:30] zul: I tried to approve the quantum in binNEW, but got: [15:30] ** quantum could not be accepted due to The following files are already published in Primary Archive for Ubuntu: [15:30] quantum-common_2012.1~e4-0ubuntu1_all.deb [15:30] gah? [15:30] nOStahl, about. potentially you could boot the installer and switch to a different vt when it comes to disk setup. But also make sure to have disbanded the raid in the bios and maybe change the mode from raid to ahci if possible in a second step [15:30] Hello, I seem to have a common problem across multiple Ubuntu Version, Multiple Pieces of hardware. Basically what happens is I do a net install via PXE boot, once it completes it boots, but stops at a flashing cursor. I know the system itself boots since I can switch to tty2 and it runs without issue. [15:31] my bios dosnt have ahci [15:31] iggi__, kinds of sounds like a video driver issue [15:31] zul: looks like python-quantumclient and quantum both provide quantum-common and they have the same version [15:32] zul: I need to reject this for now [15:32] nOStahl, ok, then at least it the disks should show up as unmanaged there. I think, bit fuzzy since every bios is a bit different [15:32] jdstrand: ok thanks [15:32] k going to reboot into ubuntu live and check out the disks there in gparted and format them [15:34] nOStahl, k, not sure how well it is done. Usually a dd if=/dev/zero of=/dev/sda bs=1M should be clean but also takes ages (with 1.5TB) [15:34] what would you say an eta would be for dd'ing [15:35] lynxman: did you got my message regarding the developers hostname hardcoded in puppetmaster-passenger? [15:36] nOStahl, no clue, sorry. can depend on a lot of things and rarely done. You may try to be smart and only erase the beginning and use skip= to erase the end in a second step [15:36] nOStahl, But then there is also the chance that the raid bios plays trick with host protected areas [15:36] patdk-lap, I've been told that before, but it's different server models. I'll dump the VGA info and check if they're the same chip, I don't think they are though. I'll just file a bug report I suppose [15:37] ugg :P [15:37] most *server* boards use that matrox 400 chip [15:37] and I believe is one of the chips with the issue [15:37] or the old ati chips [15:38] Well this one is an ASPEED technology which I've not heard of before [15:38] ASPEED Graphics Family (rev 10) [15:39] Daviey: do we need to get an FFe/release team ack for the changes we are making for rabbitmq-server? [15:39] its fixing up existing universe packages by moving them into rabbitmq-server as they can all be built from one source tree now [15:42] jamespage/lynxman: i found the lines: ext/rack/files/apache2.conf: SSLCertificateFile /etc/puppet/ssl/certs/squigley.namespace.at.pem [15:42] jamespage/lynxman: thats in the puppet source [15:42] hmm [15:42] should i create a bug for that a puppetlabs? [15:43] s/a/at [15:44] no [15:44] that gets used at the template for /etc/apache2/sites-enables/puppetmaster [15:44] jamespage: better they use example.com or whatever [15:44] jamespage: thats the developers hostname [15:45] jamespage: is there a diff for rabbit? [15:46] Daviey, linked to bug 948993 [15:46] Launchpad bug 948993 in rabbitmq-server "rabbitmq-2.7.1-0ubuntu1 lacks compatibility with plugin packages" [Undecided,Confirmed] https://launchpad.net/bugs/948993 [15:49] I have submitted an MP to fix bug 949956: https://code.launchpad.net/~benji/ubuntu/precise/lxc/bug-949956/+merge/96601 [15:49] Launchpad bug 949956 in lxc "lxc-start-ephemeral keeps same MAC address, and so same IP address, as base instance" [Undecided,New] https://launchpad.net/bugs/949956 [15:49] jamespage: so i should leave it like that ? the template? [15:49] glenn_, I'm not sure why you have the raw template - it should be re-written during configuration of puppetmaster-passenger [15:49] jamespage: that was in my vhost [15:50] jamespage: i put bla in puppet.conf [15:50] jamespage: guess its a bug then [15:51] jamespage: looking === fenris is now known as Guest71737 === Guest71737 is now known as ejat === Lcawte|Away is now known as Lcawte [15:59] Daviey: ta very much [15:59] the debdiff has had review from me and SpamapS === Ursinha is now known as Ursinha-lunch === fenris_ is now known as Guest47985 [16:30] Daviey: thanks for the reviews === swishly is now known as swish2 === swish2 is now known as swishly [16:40] jamespage: SpamapS just posted one minor correction to the debdiff [17:19] puppet/precise seems to wind up with zombie children from time to time, which it then fails to reap [17:19] puppet is weird with children anyway, they just detach it in ruby [17:19] select(0, NULL, NULL, NULL, {0, 994126}) = 0 (Timeout) [17:19] select(0, NULL, NULL, NULL, {1, 0}) = 0 (Timeout) [17:19] select(0, NULL, NULL, NULL, {1, 0}) = 0 (Timeout) [17:20] that's the parent [17:20] lamont: do you have custom modules/manifests which could cause this? I had something familiar too [17:21] dunno... a backport of 2.7.11 to lucid (and hardy for that matter) does not show this [17:21] are you on the 0.24? [17:21] 0.24? [17:21]