[00:00] ah, good to know [00:05] SpamapS: Have you talked to folks in #ubuntu-ops about it? I haven't updated a factoid for a long time, but that was helpful when I did... [00:08] I don't remember [00:10] SpamapS: what was the new factoid, i think i have foo to update it. [00:10] ? [00:11] !-ebox [00:11] Daviey: zentyal should basically be the same as ebox with the string replaced, and ebox should point to the zentyal factoid [00:11] I already ported the Zentyal page too https://help.ubuntu.com/community/Zentyal [00:12] !servergui probably wants to be updated to match too === robbiew is now known as robbiew_ [00:13] ebox has no aliases - added by LjL on 2008-01-16 23:51:42 - last edited by jussi01 on 2008-04-25 19:25:35 [00:13] shauno: I am only a bot, please don't think I'm intelligent :) [00:17] !zentyal [00:17] What did you do to poor ubottu ? [00:17] !ebox [00:17] ebox is a web-based GUI interface for administering a server. It is designed to work with Ubuntu/Debian style configuration management. See https://help.ubuntu.com/community/eBox [00:17] !zentyal [00:17] zentyal is a web-based GUI interface for administering a server. It is designed to work with Ubuntu/Debian style configuration management. See https://help.ubuntu.com/community/Zentyal [00:17] \o/ [00:21] !ebox [00:21] zentyal is a web-based GUI interface for administering a server. It is designed to work with Ubuntu/Debian style configuration management. See https://help.ubuntu.com/community/Zentyal [00:22] Daviey: ty! [00:22] yay for zentyal :) [00:23] SpamapS, Wonder if it should reference that it was formally known as ebox? [00:25] !servergui [00:25] Ubuntu server does not install a desktop environment or X11 by default in order to enhance security, efficiency and performance. !eBox provides a GUI system management option via a web interface. See https://help.ubuntu.com/community/ServerGUI for more background and options. [00:25] altho it's still ebox in LTS, so whether that's wrong or not is possibly debatable [00:27] shauno: its mentioned in the linked page [00:28] !ebox [00:28] zentyal is a web-based GUI interface for administering a server. It is designed to work with Ubuntu/Debian style configuration management. See https://help.ubuntu.com/community/Zentyal (Project formally known as eBox - including in Lucid/10.04). [00:28] Everyone is happy :) [00:28] Daviey: no, I'm pretty sure smoser is still a little peeved. [00:29] * Daviey hopes he is :) [00:32] So, I just heard about zentyl: is my brief reading correct that this is a web-accessible administration framework perfect for use on a small NAS box? [00:33] persia: Anytime you put framework and perfect in the same sentence, you are certain to be wrong. [00:33] This is part of why I'm asking :) [00:33] Then your answer is "No." [00:34] I can tell that without knowing anything about zentyl. [00:34] All the past remote administration tools I've heard about have grave issues of one sort or another. [00:34] It may suggest a different question is in order. [00:34] Indeed. Perhaps I'm asking for some hints as to how it isn't perfect. [00:34] Agreed. The only non-broken remote administrative interface I'm aware of is ssh. [00:34] And it's not perfect. [00:35] Since it doesn't appear to be packaged, I expect it's unlikely you'll find much experience with it here. [00:36] According to https://help.ubuntu.com/community/Zentyal it's a rename of ebox: I figured the rename just hadn't hit oneric yet. [00:38] First I heard of the rename. [00:39] We did do some work with ebox devs several cycles ago, but it seems to have died out. [00:39] Hrm. Oh well. Based on the impetus information that led to me being interested, I figured folk who are often here would have been involved, but if it's not been discussed here (obviously, or you'd know), and as PPAs are being recommended (always indicative of questionable practices), I suppose I'm incorrect. [00:40] It's been awhile, but the best thing I remember they gave up on being current in Ubuntu and just went to a PPA. [00:40] It was better than webmin, but not particularly wonderful as I recall. [00:40] better than webmin isn't saying much :) [00:41] I'm not saying it's bad either. That was awhile ago. [00:42] Understood: I'm interpreting your comments as "I'm not that familiar, but I'm trying to be helpful" [00:43] Maybe Daviey or SpamapS will catch backscroll, and have recommendations (or at least some understanding why the Ubuntu documentation is recommending a PPA) [00:44] That and "this laptop is too heavy to carry open, one handed while I try to navigate to the table without losing my wifi connection at the restaurant." [00:44] You should know better than to thing that anything in the community section qualifies as Ubuntu documentation. [00:44] And this is why one shouldn't bother brining a laptop > 600g to a restauran [00:44] t [00:45] When I left the house I didn't know I was going to a restaurant. [00:46] And I'll be the first to admit that while one can do a lot of things on a super light platform, thre are limitations [00:47] the ebox/zentyal guys were at UDS-O [00:47] hello, is there a correct way to get a more up to date apache2/php setup on 10.04 ?? [00:48] i looked in backports but did not see anything [00:48] SpamapS: Will zentyal be coming to oneiric? What are some of the disadvantages? [00:49] ebox == zentyal in oneiric [00:49] they just haven't renamed the packages [00:50] upstream do not put as much effort into non-lts releases.. but they are starting to ramp up for 12.04. [00:50] Which is unfortunate.. we should probably help them do that. [00:51] well, aiui, they consider it a 'best effort'... but not supported by upstream [00:52] That's kinda unfortunate, as it makes transitions a lot harder. [00:53] Pretty slick screencast [00:53] I'd hope they'd have a flexible enough build system to do the right thing on all of development, current, and current LTS. I don't think it's an issue if they only offer end-user support for LTS, but developer support should be wider. [00:54] persia, yeah... it's best effort... I can't argue too much with that [00:55] I can't argue with the wording, but lots depends on what it means. If it means "We'll work on it, but don't bother us too much unless you're prepared to help a bit", that's fine. If it means "We might get around to that if we have time", that's less exciting. [00:58] persia, they did a pretty complete lecture at FOSDEM this year, I think the video went online.. looking [01:01] bah, can't find it [01:02] One of the honcho's does frequent here, i'll push him your way if you have further questions. [01:03] Thanks. === rickspencer3_ is now known as rickspencer3 [02:17] anyone using a Drobo (specifically Drobo-s) with server? any tips on formatting, lun size, etc? [02:25] willl do-release-upgrade upgrade my LTS to the latest LTS or will it go to 10.10 [02:26] sabgenton: it can do both. check /etc/update-manager/release-upgrades for a line reading Prompt= [02:26] as in 10.04 to 10.04.2 [02:26] is what I want [02:26] ah [02:26] can i do that? [02:26] sabgenton, For that, `apt-get dist-upgrade` is sufficient. [02:26] That is, if you have -updates enabled in your /etc/apt/sources.list [02:27] whats the difference between dis-upgrade and upgrade [02:28] sabgenton, It's slightly different handling of cases where packages conflict or replace each other. [02:28] when do I use upgrade [02:28] The apt-get(8) manpage describes the differences in fair detail [02:28] k [02:28] ok dist-upgrade it is [02:29] Unless someone had to do something fairly strange, 10.04->10.04.2 should be OK with just "upgrade". I tend to use dist-upgrade when I *know* I want to upgrade just to work around things. [02:29] so does do-release-upgrade upgrade do what I want though? [02:29] or will it allways up the version number to the current stable [02:29] just currious [02:29] No. do-release-upgrade will try to move 10.04 to 10.10 or 12.04 (depending on configuration). [02:30] ok so u can choose verison numbers but not the same version number [02:30] Rather, you can choose between "next 6-monthly release" and "next biannual release". [02:31] "when I *know* I want to upgrade just to work around things." [02:31] eg? [02:31] Sometimes there's an upgrade of something where the developers need some package to be uninstalled or a different package to be installed. For stable releases, this is extremely rare. [02:34] so dist-upgrade fixes dep isuse where as upgrade just upgrades? [02:34] or fixes more dep isuuses [02:35] at least [02:35] Treats dep issues slightly differently. dist-upgrade is allowed to uninstall something to resolve things (it will tell you what). upgrade is not. === dead_letter is now known as MTecknology === hallyn is now known as hallyn_afk [03:36] SpamapS: 'https://wiki.ubuntu.com/MeetingLogs/Server/20110405' [11:09:25] SpamapS : to put his version of ubuntuserver-minutes in directions for writing minutes [03:42] persia: thx for before :) [03:54] http://paste.debian.net/117892/ is it just me, or does NFS root_squash break setgid? [04:02] New bug: #787891 in openvpn (main) "openvpn should soft-restart on interface up" [Undecided,New] https://launchpad.net/bugs/787891 [04:37] i have to pick and build a 2.6.36+ kernel for a hyper-v guest (for the staging drivers). any advice on particularly good/bad versions? [04:44] hyper-v can't emulate a 486? [04:45] dunno, what does it matter though? [04:47] Well, .32 has drivers for a 486 [04:49] does it have the ethernet accel drivers? i want to avoid using the legacy hyper-v adapter mainly [04:49] I imagine that "accel" means "not a 486" [04:50] the equivalent of linux KVM's virtio stack [04:50] exactly [04:50] accel in the "virtio hyper-v equivalent" sense [04:51] So what I'm saying is: unless you demonstrably need that extra I/O throughput, don't try to fuck with your distro's stable, tested kernel === nigelb_ is now known as nigelb [04:55] allright [04:55] i know they are in staging, but maybe they are working well [04:56] s/they/the hyper-v drivers/ [05:38] hi folks... can someone help me in configuring openvpn? i want to ensure that *only* a few types of traffic go through the vpn. i.e. like web browsing. right now i'm using hidemyass's openvpn config and it seems to be sending *everything* through the vpn. [05:45] hallyn_afk: its in the comments of wiki.ubuntu.com/ServerTeam/Meeting === smb` is now known as smb [08:02] I just set up a new raid1, and the new one has "super 1.2" in /proc/mdstat in addition to my old md0 which doesn't. Anyone knows what this "super 1.2" actually mean? [08:17] speakman: its telling you that the format of the super block is the new 1.2 format [08:31] oh, so theres a new format.. *googling* :D [08:53] speakman: sounds like the superblock is at /dev/sda1 [08:53] assuming retarded grub2-style numbering [09:03] Can one "upgrade" the superblock easily? [09:05] by the way - which filesystem is to prefer for file storage these days? [09:10] speakman: ignore me, I'm talking rubbish [09:10] As to filesystem -- btrfs as soon as it's ready, and ext[432] until then [09:14] twb: Ok. I'll just keep on with my extfs then. :) [10:31] New bug: #788004 in postfix (main) "package postfix 2.8.2-1ubuntu2.1 failed to install/upgrade: subprocess installed post-installation script returned error exit status 75" [Undecided,New] https://launchpad.net/bugs/788004 [10:31] New bug: #788006 in samba (main) "package winbind 2:3.5.8~dfsg-1ubuntu2.2 failed to install/upgrade: subprocess installed post-installation script returned error exit status 1" [Undecided,New] https://launchpad.net/bugs/788006 [10:38] hi, i'm on a Lucid server and i ry to use openssl with IPv6... but i cannot do a s_client -connect IPv6 --> i get connect: Invalid argument... do you know a solution? [11:16] Hello guys, could you please help me with the partitioning thing! it is making me crazy and i can't understand it. I have two hard discs each one is 1tb and I am intending to install zimbra email server on my 10.4.2 server. thnx [11:20] :( [11:22] What are you trying to do with the two disks? [11:26] persia I want to install Zimbra email server and the seller told me that buying tow hard discs is better for a server! i know how to manage to email server but I don't know about portioning. [11:26] partitioning * [11:29] Right. So there's a number of ways that two drives can be an advantage. [11:30] 1) you can put the OS on one and the data on the other [11:30] 2) you can mirror the data between them for reliability [11:30] 3) You can charge more when selling a server and get a larger commission [11:30] yes. [11:30] So, you have to decide which of those applies in your case. [11:31] I want to mirror the data between hard discs for reliability. [11:32] https://help.ubuntu.com/10.04/serverguide/C/advanced-installation.html : read the section on "Software RAID". [11:33] (the procedures are unchanged from 10.04 to 10.04.2) [11:38] Syria: http://www.cybersprocket.com/2010/blog/setting-up-raid-1-on-ubuntu-10-04/ [11:39] lambda_x: Thx [11:40] np [11:40] can I add sources by adding files to /etc/apt/sources.list.d or is it for something else? [11:41] "Select the first hard drive," the 1st hard drive is SCSI3 (0.0.0) (sda) ? [11:41] Syria: you got sda and sdb I guess?, so first will be sda [12:17] lambda_x: Still here? [12:22] Hi, someone already deals with openssl s_client connect? [13:20] hallyn_afk: morning, poke me when you're around [13:46] New bug: #788091 in samba (main) "package smbclient 2:3.5.8~dfsg-1ubuntu2.1 failed to install/upgrade: corrupted filesystem tarfile - corrupted package archive" [Undecided,New] https://launchpad.net/bugs/788091 [14:00] hey guys how do i create another session with byobu? when i just type byobu it reattaches to the current current session. I want to start a new session with it [14:00] Doonz: byobu -S new [14:01] Doonz: just launch, but give the new one a different title [14:01] Doonz: or, you can do "byobu bash" if you don't care about the title [14:01] Thank youi wasnt using the -S switch and it was comaplaining about the directory [14:01] Doonz: no problem, enjoy ;-) [14:02] thanx again [14:05] kirkland: would it be better if we linked the IRC channel on the top of the room name and possibly in all the session pages? [14:05] kirkland: um, re:summit :) [14:05] nigelb: yes! [14:05] nigelb: i'm sorry my code sucked :-) [14:05] nigelb: but I think it's a good idea [14:06] kirkland: heh, that was in the works last time (I believe I still own the bug), will try to get it out for next time :) [14:06] nigelb: it would be *awesome* to 1) install chatzilla as a firefox plugin, 2) click on irc link and land in channel [14:06] nigelb: sweet, you da man! [14:07] :) [14:18] kirkland: whats up? [14:20] hallyn_afk, shouldn't you be afk? === hallyn_afk is now known as hallyn [14:21] Daviey: nyeh [14:22] lynxman: rabbitmq-erlang-client rabbitmq-stomp needs to be updated to 2.4.1 in order for it to work in oneiric [14:23] zul: library dependencies? [14:23] zul: shouldn't be a problem I reckon [14:23] lynxman: nope rabbitmq-server is now 2.4.1 on oneiric [14:23] zul: oh darn *snap* [14:23] zul: okay, putting that in my todo :) [14:23] lynxman: thats why mcollective is having problems connecting to the stomp server [14:24] zul: yeah plugins are very version dependant [14:25] zul: I'll redo the packages for 2.4.1, hopefully today, refactoring the provisioner code *sob* [14:29] Daviey: so for reviewing blueprints do we put the comments at the bottom? [14:30] zul, Well... smoser raised a valuable point that it doesn't store history there. [14:31] I did think about sending a mail to the server list with the title of each bp [14:31] *but*, i think i'd be criticised for noise. [14:31] So.. probably better to stick with the whiteboard. [14:31] Daviey: sure but if you are subscribed to the spec then you get a diff of the whiteboard change [14:31] exactly [14:40] zul: is there anyway we can liaise the rabbitmq plugin packages to rabbitmq, so there's some sort of alert or dependency trigger... or the same maintainer [14:41] lynxman: i usually take care of rabbitmq-server but i wasnt aware of rabbitmq-stomp and erlang-client when i did the merge ill be more careful next time [14:41] zul: aah okay :) [14:44] zul: I'll get you the new packages in some mins [14:45] Daviey: o/ so you ended up not emailing anything after all, not even on how to proceed on commenting the BP;s? [14:54] lynxman: pardon my ignorance :-) ... but ./usr/share/orchestra/rsyslog/orchestra-client.conf:*.* @@1.1.1.1:514 [14:55] lynxman: what's that in rsyslog parlance? [15:02] RoAkSoAx, at the moment, no. [15:04] lynxman: nevermind, I think I'm digesting it now [15:04] Daviey: k [15:05] kirkland: hah sorry, was in PPA hell [15:05] kirkland: let me know if I can help :) [15:06] lynxman: it's okay; i'd like to get an orchestra 1.0-ish into archive review ASAP [15:06] kirkland: yeah I know, I would feel better if I could integrate nagios beforehand, but... it's in a stable place now [15:06] lynxman: nah, we need to cut this version loose and keep rolling [15:06] kirkland: rolling rolling rolling... rawhiiiide [15:07] lynxman: your knowledge of Americana never ceases to amaze me :-) [15:08] kirkland: Well my mother is from Panama... so it's almost as if I've been raised in Miami ;) [15:08] lynxman: heh [15:10] it's almost the same, except for the corrupt government and the poor areas surrounding the city [15:10] so it's the same [15:10] :D [15:10] ... [15:12] zul: https://launchpad.net/~lynxman/+archive/ppa [15:12] err nevermind, build error [15:16] build packages in the wrong order, shame on me :) [15:27] lynxman: http://paste.ubuntu.com/612759/ [15:27] lynxman: do those permissions look right? [15:27] lynxman: that's a lot of 600 ... are there passwords or private data in those files? [15:28] kirkland: in the seed one there is, and cobbler runs as root so it should be okay I imagine [15:30] lynxman: one more thing ... debian/ubuntu-orchestra-monitoring-server.install:monitoring-server/* / [15:30] lynxman: there's nothing in that dir yet [15:30] lynxman: i'm going to drop that install file for now [15:30] lynxman: add it back when you get the nagios bits [15:30] kirkland: sounds good [15:31] New bug: #788146 in bacula (main) "package bacula-director-mysql 5.0.3-0ubuntu2 failed to install/upgrade: sub-processo script post-installation instalado retornou estado de saĆ­da de erro 1" [Undecided,New] https://launchpad.net/bugs/788146 [15:53] SpamapS: are you around? [15:57] wtf - just upgraded a VM here from 10.10 to 11.04, and it starts up with just grub telling me it can't boot [15:57] any ideas how to fix that? [15:57] * RoyK doesn't like grub2 === NG_ is now known as ng_ [16:25] hi [16:25] hi [16:25] been running a software raid5 in debian for quite a while - it's very stable, no problem at all so far. how about ubuntu server here? [16:26] DexterF: running raid5 in my ubuntu server, no issues [16:26] Dont think you will see an issue. I run zfs and its been rock solid. [16:27] aurigus: zfs via fuse? [16:27] until a week ago. [16:27] ..and now? [16:27] The posix layer has just been released for zfs natively on linux [16:28] So now I'm running natively :) [16:28] how come this wasnt in the newstickers? [16:28] good news, but how were the licensing issues resolved? [16:29] I was going to make a big deal about it, blog posts and such, but it is still beta [16:29] RC4 [16:29] I wasn't sure of its stability either... but for the last week or 2 its been rock solid [16:29] performance over linux-md? [16:30] Haven't had time to do any benchmarking [16:31] hallyn: pong.. whats up? [16:32] reason I'm asking: I'm planning to consolidate file server and htpc. now debian stable is not really the best platform for htpc and testing not for a raid fileserver [16:32] hence I'm eyeing 10.04 [16:33] * patdk-wk just uses a seperate box for both of those [16:33] I'm running it on 11.04 and its working fine [16:33] Im just using it as a file server though [16:35] SpamapS: i was just remining you about an action you had last month, to put up your meeting notes posting script in the knowledgebase :) [16:35] hallyn: right, I guess in the comments of the meeting page isn't all that discoverable. ;) [16:35] there used to be a "how to do the minutes" page [16:36] lol [16:36] well i always look at https://wiki.ubuntu.com/ServerTeam/KnowledgeBase at the bottom [16:38] ahh thats the page [16:38] stupid moin's sucky search [16:41] hallyn: I updated the link to point at mine, since it supersedes mathiaz' old branch [16:48] SpamapS: cool, thanks [16:49] http://karlsbakk.net/grub.png <-- this is all I get after upgrading this VM to 11.04 - any idea what to do next? [16:57] Im curious if anyone can give me some advice regarding RAID arrays [16:58] shoot [16:58] Right now, I've got a 4TB file-syste, (3x 2TB disks) in a RAID5. and I've ordered two new 2tb disks to expand as Im running out of space. Would I be better off upgrading to RAID6 and giving me 6TB of storage or would RAID5 be safe-enough with 8tb [16:59] I do weekly backups of all my files as is, (although starting to run out of backup space haha..) [17:00] raid-5 is rather safe, but new drives tend to be worse than older drives due to density [17:00] Slyboots: given that 1TB recoveries will probably take *hours* .. I would go w/ RAID6 [17:00] Mmm.. [17:00] err.. I read the original wrong.. 2TB .. RAID5 would be almost insane. [17:01] and I'm not sure if you can change from raid-5 to raid-6 without recreating the raidset [17:01] Slyboots: did you test how long a re-sync takes when you created the RAID5 ? [17:01] SpamapS: haha.. it takes.. quite some time [17:01] Yeah thats a backup / restore operation [17:01] about 8-9 hours [17:01] SpamapS: insane? [17:01] 8-9 hour window for another drive to fail... add on the time between failure and replacement.. [17:01] Wow, well.. wiping the server will be a major pain in teh ass [17:02] seems changing it should be possible http://neil.brown.name/blog/20090817000931 [17:02] Slyboots: honestly this is why you're better off with 10 little drives than 5 giant drives. [17:03] RoyK: nice! [17:03] how can I make "hostname" give "example.com" instead of "example" ? [17:03] Well Im in the process of backing up my server. [17:03] Im actually tempted to shfit the OS onto its own disk.. and just not RAID it at all [17:03] bencc: the hostname should be a single word - the domain is another thing [17:03] bencc: see /etc/defaultdomain for the domainname [17:04] bencc: hostname -f [17:05] RoyK: I think in this case he wants his FQDN to be example.com .. not just the domain [17:05] bencc: you'll probably want to look at changinging /etc/hostname [17:05] but why? [17:06] Because he wants his website to be that server? [17:06] Its a perfectly valid configuration and quite common. [17:06] SpamapS: no need to change the hostname for that - simply a matter of configuring dns+apache [17:07] I mean, *I* wouldn't do it, but its not a huge problem. [17:07] I need several application to get the "same" hostname [17:07] puppet and erlang [17:07] one thinks it's example and the other example.com [17:07] imho it's rather silly - adding an A record to allow for http://example.com/ isn't hard, but setting the hostname to the FQDN isn't really a good idea [17:09] So I just add the new disks to the RAID array. then run mdadm --grow ... [17:09] And that'll conver it to RAID6 [17:09] RoyK: why not? [17:10] because one day you want another host in that domain [17:10] erlang is particularly picky about hostname [17:10] and the normal way to set things up, is to set a hostname on the host and add that hostname to a domain [17:11] SpamapS: how does it get inet:gethostname() ? [17:11] bencc: to RoyK's point, it would be better to pick something like 'web00.example.com' and let example.com be just a DNS alias. [17:11] a programming language shouldn't really bother with the hostname on which it's running [17:11] SpamapS: not all DNS servers allow for FQDN to be registered as a CNAME record - but then - adding another A record won't hurt [17:12] RoyK: unless it is distributed... [17:12] bencc: at boot up, /etc/hostname is read and fed into the 'hostname' program. So you can change that file, and then run 'sudo hostname `cat /etc/hostname`' and then restart any services you need to see that hostname. [17:12] RoyK: whoa whoa nobody said anything about a CNAME [17:12] SpamapS: hostname -F /etc/hostname [17:12] alias != CNAME [17:12] kirkland: you about? [17:12] SpamapS: same thing [17:12] RoyK: man pages FTW ;) [17:12] nigelb: howdy [17:12] hostname -F /etc/hostname doesn't effect erlang inet:gethostname() even after reboot [17:13] kirkland: novacut folks have a vid of you on the bass :) [17:13] * pmatulis loves his bandwidth (wget output: 713,385,984 9.98M/s in 97s) [17:13] RoyK: no, a CNAME is most certainly *not* the same thing. CNAME is for giving away control of one hostname to another domain. [17:13] SpamapS: no, it certainly is not [17:13] ok, whats it for then? [17:14] it's an alias [17:14] Common NAME [17:14] nigelb: hehe :-) yikes [17:14] SpamapS: delegation isn't done in the zonefile - it's done in the bind config (or whatever dns server you use) [17:14] heh [17:14] Canonical name actually [17:14] well, still, it doesn't have anything to do with delegations [17:15] CNAME == alias [17:16] I didn't say delegation, thats a word with a lot of meaning in DNS. I used small words because its a conceptual definition. You give away control of a single hostname to another domain not in the same zone.. if it were in the same zone, you could just set the address to the same thing. [17:16] Some people certainly misuse it that way. [17:16] But they're wasting their DNS server's time and resolvers' time doing two fetches. :-P [17:17] Anyway, they are aliases, yes, but they are not the only way to make an alias. [17:18] And are probably the worst way. [17:18] Need help, trying to install ubuntu 64-bit server ver 11.04, i can choose language and when i choose install ubuntu server it just reboots. no error no nothing, any ideas? [17:19] D0minat0r: installing from a CD? [17:20] SpamapS: yes [17:20] the dvd-rom spinns to speed then reboot [17:21] D0minat0r: are you certain you have a 64-bit machine? [17:22] yes and i have tried 32bit installer too same thing [17:22] D0minat0r: very strange indeed. [17:22] D0minat0r: I believe there is a memory test option on the CD.. maybe you could run that? [17:23] running [17:26] SpamapS: takes a while... :) [17:30] D0minat0r: indeed, though I'd say if it runs through the first 2 passes you can cancel it and we can move on to other diagnostics [17:30] it passes no errors [17:32] can someone shed some light on bridge interfaces? I have a couple setup but can't pass any traffic [17:32] i read that it needs cap_net_admin capability due to a kernel bug. Is this still true ? [17:35] peydude: I have a server running kvm with bridging - works well [17:36] that's exactly what I am trying to accomplish. I have two guest servers, my host can pass traffic fine (with or without vlan tagging) but when my bridges are setup [17:36] the guest OSs are not accessible [17:38] peydude: pastebin your network config [17:38] as in /etc/network/interfaces [17:39] as well as your guest xml files [17:41] SpamapS: i took away 2 gig ram of 4gig and it works [17:41] dual channel memory [17:42] D0minat0r: hm weird [17:42] Here is the host interface file: http://paste.ubuntu.com/612826/ [17:42] pretty basic [17:42] let me grab my guest xml [17:43] peydude: virsh dumpxml name > name.xml [17:46] guest xml: http://paste.ubuntu.com/612827/ [17:48] peydude: your bridge config is unusual. i usually use just a single line but you have 12 [17:48] peydude: does br0 even work? [17:49] pmatulis: you are right. I have specified how the system should bring the bridge down and up to avoid any errors (e.g. bridge already exists ... ) [17:49] pmatulis: no. if I just attach a single interface to it it won't pass any traffic regardless of it being a kvm interface or a host interface [17:50] pmatulis: that's what i was asking about bridged interfaces to begin with. [17:50] why* [17:51] peydude: do you have bridge-utils installed? [17:51] pmatulis: you are probably used to a setup like the one in this doc: https://help.ubuntu.com/community/KVM/Networking [17:51] pmatulis: yep otherwise i couldn't have setup the bridge :) [17:52] peydude: but you said it doesn't work [17:52] peydude: that's pretty vague [17:53] peydude: i recommend setting it up manually [17:53] pmatulis: let me set it up manually and see if I can get any traffic to go through ... brb [17:54] I set postfix dovecot w/ sieve and it is work, but to make it work I remove maildrop now my mail does not get delivered [17:54] Jcook_5xData: why did you need to remove maildrop to make it work? [17:55] weel I guess I dint not but maildrop does not read .dove.seive file does it has it own filter [17:56] I am doing all this because of roundcube does not support filter other now sieve [17:57] Jcook_5xData: put maildrop back. you may then need to edit master.cf (comment maildrop line out) [17:58] D0minat0r: I'ts probably worth reporting a bug about it when you've got the info settled down, (also interesting to see if it works if you re-add the memory post-install for example..) [18:01] done [18:09] pmatulis, do I need a deliver agent then? [18:09] I can not find any good how to on this === hallyn is now known as hallyn_afk === The is now known as Guest48296 [18:39] arand: well i think it has something motherboard because the system wont function with the 2 onboard LAN devices, when i activate them i have to take aay more ram and only have 1gig of ram for it to run [18:39] its so strange right no [18:41] D0minat0r: Yes, very odd [18:42] motherboard is Abit in9 32-MAX [18:42] crap [18:44] * RoyK just came in - D0minat0r what's up? [18:45] RoyK: trouble getting ubunut to run on that mobo [18:46] D0minat0r: what does it say? [18:47] a sec, re installing [18:47] had to take away 1 gig ram for the installer to work with the onboard lan devices [18:47] wow its been years since I cared about motherboards.. ahh the good old days [18:48] with 4 gig the installer just reboots, with 2 gis i have to have onboard lan devices disabled and with them enabled i only can have 1 gig ram :O [18:48] D0minat0r: seems something is rather fscked up [18:48] you think? :D [18:49] wtf now, disk boot failure [18:49] and no errors during installation [18:53] D0minat0r: RMA the m/b? [18:55] pmatulis: what is that? [18:55] return it [18:55] hehe [18:55] too old [18:55] freadking crap tho [18:55] then perhaps get a new one :P [18:56] you get el-cheapo atom-based mobos for almost nothing [18:56] including the cpu and perhaps some memory [18:57] http://cgi.ebay.com/FOXCONN-45CS-W-INTEL-ATOM-230-SATA-DDR2-LAN-MOTHERBOARD-/250812797777?pt=Motherboards&hash=item3a659b9351 [18:58] pmatulis, Ok I have mail getting delivered without maildrop, for some reason it is still not obeying the .dovecot.sieve file. do you know any go how to or have you done this and can give me pointers [19:05] Jcook_5xData: no, never used it. should soon though [19:09] not any good how out there... well that I can find [19:09] pmatulis: i am able to get bridge networking to work with the guest interface(s) attached to my first host bridge [19:10] pmatulis: but as soon as i attach a second guest interface to the a second host bridge the guest becomes inaccessible [19:10] still running one guest OS [19:11] peydude: you can use a single br0 for all host and all guest activities [19:12] zul: howdy! since you merged openldap, you might have an idea why ipsec-tools FTBFS when building in Oneiric against openldap. But, when building in natty it does not fail: https://launchpadlibrarian.net/72308605/buildlog_ubuntu-oneiric-amd64.ipsec-tools_1%3A0.8.0-3ubuntu1~ppa1_FAILEDTOBUILD.txt.gz [19:12] pmatulis: yes i have attached two guest interfaces (from the same guest OS) to br0 and I can still access the guest [19:12] peydude: well, you can do the same for all guests [19:12] pmatulis: the problem comes when I attache the second guest interface to another bridge on the host [19:13] peydude: why another bridge?? [19:13] pmatulis: vlans! what if I want to attach the second guest interface to a host bridge that is on a different vlan [19:14] peydude: ok [19:14] pmatulis: the problem might be the mac address of the second bridge (since it's running on the same physical interface on the host) === CyrusB_1 is now known as CyrusB [19:15] peydude: note that i didn't see a 2nd bridge configured in your interfaces file === ng_ is now known as NG_ [19:17] pmatulis: correct. I added one manually :) i am not blaming you :) [19:20] peydude: well, the manual one is working? [19:21] pmatulis: only with the setup i described above (all guest interfaces attached to the first bridge on the host interface). [19:22] peydude: so, remove the auto one and add it manually too [19:22] pmatulis: I am testing it with a second bridge on the host interface now by changing the mac address of the second bridge [19:23] zul: I'm also merging facter and it seems that in natty you've carried merges in debian/rules that are not represented in the changelog... http://paste.ubuntu.com/612878/ [19:23] smoser: why do 64bit amis have no swap, but 32bit ones do? [19:23] nor there's a convincent explaination why they have been carried over [19:23] or is it *.large vs *.small? [19:24] zul: should I drop them? [19:25] RoAkSoAx: on a call gimme a few [19:25] (note that part of those changes seems to be when cdbs was dropped in ubuntu and twhen it happen in debian, they didn't update debian/rules, though the rest of the delta can be dropped for what I can see [19:25] zul: sure [19:26] RoAkSoAx: ergh....yeah dont drop them [19:27] zul: k, I'll just drop the stuff in configure-stamp as that's just lines that makes no sense carrying over then [19:28] RoAkSoAx: k thanks for the ldap stuff thats new to me [19:29] zul: yeah being banging my head yesterday trying to figure out why ipsec-tools won't build with the newer ldap [19:29] RoAkSoAx: ill take a look [19:29] zul: k ;) [19:30] thanks [19:40] jiboumans, that comes from amazon [19:40] smoser: figured =/ [19:40] smoser: would it make sense to use the values you put in for m1.small/medium on the large ones [19:40] you can partition a ephemeral disk and add swap [19:40] yup [19:40] or just use a file on /mnt [19:41] right. [19:41] riak and rabbitmq are both very very unhappy when they run out of ram [19:41] so it seems prudent to add some overflow and make the monitors scream bloody murder [19:42] sudo apt-get install bind9 gives me * Starting domain name service... bind9 [fail] [19:43] Anyone can help? [19:47] Someone msg me in this stupid client I cant see msg :/ [19:47] post: see ya [19:53] syslog syslog where are you === pangrazi1 is now known as pangrazi [20:00] post: running? [20:00] RoyK: No still getting the error and I cant find syslog :( [20:00] ps axf [20:00] does that list syslogd? [20:04] RoyK: it listed something [20:06] are these the steps i must go through to create a self signed x509 cert? https://help.ubuntu.com/community/OpenSSL [20:08] RoyK: nothing that even looks like syslogd (or bind9) is listed [20:12] Slyboots: http://www.snia.org/education/tutorials/2008/spring/storage/Whittington-W_Desktop_Nearline_Enterprise_HDDS.pdf <- read the section about UER propability when recovering, then do the math for your setup. (you'll end up with: you want raid6. and: you want to rebuild the array and backup/restore, not expand it. and: you want, let's say a HP P400 Smart Array with BBU) [20:15] WEll I dont want to rebuild the array if I can help it [20:17] I don't see how raid5->6 could even work. I mean it's possible but I so would not want an array that has a pattern originated from a rather obscure stunt but something fresh from scratch. we're talking data integrity here. [20:24] Well mdadm is supposed to support it [20:25] officially? wow, they've come a long way. [20:26] trying to generate my first x509 cert: https://help.ubuntu.com/community/OpenSSL ... but curious when it says "Create the server configuration file, by editing ~/myCA/exampleserver.cnf with your favorite text editor. Add this example content:" ... there is no exampleserver.cnf generated prior to that step [20:27] New bug: #788284 in facter (main) "Please merge facter 1.5.9-1 (main) from debian unstable (main)" [Wishlist,In progress] https://launchpad.net/bugs/788284 [20:30] RoAkSoAx: is facter done yet? :) [20:30] anyone ever used checkinstall? why ubuntu tells me it cant find the bin that is installed with checkinstall === Tohuw is now known as Guest90795 [20:33] Daviey: give me a sec [20:33] I'm uploading debdiff [20:33] Daviey: suit yourself [20:40] pmatulis: I got it working :) [20:41] peydude: please do tell [20:41] pmatulis: I added vlan interfaces to all the host bridges (aka tagged vlans). this way when my guest interfaces attach to the bridges they all work [20:42] pmatulis: before I had untagged vlan traffic on eth1 and tagged on eth1.5, eth1.200, etc [20:42] pmatulis: that didn't work. so i stopped using eth1 and just utilized its vlan sub-interfaces [20:42] peydude: fair enough [20:43] peydude: maybe post your new interfaces file [20:44] pmatulis: i have done everything manually so far. I will update my files and post them [20:44] negronjl: ping [20:45] RoAkSoAx: hi [20:45] negronjl: o/ hey, I was wondeirng if you have a howto on how to test https://launchpad.net/bugs/732953 [20:45] Launchpad bug 732953 in facter "can_connect function inside ec2.rb always return false" [Low,Fix released] [20:47] RoAkSoAx: let me check really quick...hold on [20:47] RoAkSoAx: withouth the fix. run facter ec2_ami_id or facter | grep ec2 [20:48] RoAkSoAx: nothing should come up [20:48] RoAkSoAx: with the fix: [20:48] RoAkSoAx: run the same and something like this: [20:48] $ec2_ami_id [20:48] $ec2_ami_launch_index [20:48] $ec2_ami_manifest_path [20:48] $ec2_block_device_mapping_ami [20:48] $ec2_block_device_mapping_ephemeral0 [20:48] $ec2_block_device_mapping_root [20:48] $ec2_block_device_mapping_swap [20:48] $ec2_hostname [20:48] $ec2_instance_id [20:48] $ec2_instance_type [20:48] $ec2_kernel_id [20:48] $ec2_local_hostname [20:48] $ec2_local_ipv4 [20:48] $ec2_placement_availability_zone [20:48] $ec2_public_hostname [20:48] $ec2_public_ipv4 [20:48] $ec2_public_keys_0_openssh_key [20:49] $ec2_reservation_id [20:49] $ec2_security_groups [20:49] should show up [20:49] I'll be back in a minute [20:51] negronjl: that's from an instances? or from a machine with ec2 tols configured or something like that? [21:01] !pastbin @ negronjl [21:01] RoyK: I am only a bot, please don't think I'm intelligent :) [21:01] !pastbin [21:01] !paste [21:01] For posting multi-line texts into the channel, please use http://paste.ubuntu.com | To post !screenshots use http://imagebin.org/?page=add | !pastebinit to paste directly from command line | Make sure you give us the URL for your paste - see also the channel topic. [21:01] !pastabowl | IdleOne [21:01] hehe [21:02] RoyK: !pastebin is already an alias to [21:02] this however, isn't the correct spelling of anything; !pastbin [21:02] guess my slight typo was the issue :P [21:02] :) [21:03] pastbin - whatever happened five months ago? [21:05] Daviey: i'll let you know when facter is done as I'm waiting for someone to test a patch that might be able to drop [21:06] RoAkSoAx, I'm actually build testing your debdiff atm [21:07] RoAkSoAx, Should i hold off uploading it then? === hallyn_afk is now known as hallyn [21:07] Daviey: that works, but was further looking into it and the patch to ec2.rb might be able to be dropped as it seems the error was in a different place and has been addressed differently [21:07] Daviey: yes pelase, just waiting for negronjl to give it a test the way he is used to, and wanna see the results [21:07] k [21:08] Daviey: it is not really gonna hurt if it gets uploaded now, though, I'd prefer to not do a second upload tomorrow >P [21:08] Daviey: but if you wanna sponsor somethign for me corosync, pacemaker, and cluster-glue are still in the queue :) [21:09] RoAkSoAx, yeah, if it was a case of "not sure when i'll have the patch" - i'd just upload it, but as you are activly working on it - lets wait :) [21:09] * Daviey wonders when RoAkSoAx will apply for access to upload this stuff himself. [21:10] Daviey: i';ll do this week [21:10] Daviey: was catching up with other sutff first [21:10] :) [21:14] RoAkSoAx, bah, "mark as patch" [21:16] http://karlsbakk.net/fun/swineflupooh.jpg [21:17] when I log into lucid server I see welcome msg twice: for 10.04.2 and 10.04 [21:17] is this a bug or is it normal? [21:17] ChmEarl: When did you install? [21:17] it's the motd.tail bug [21:18] let me see.. [21:18] just remove /etc/motd.tail [21:18] Daviey, installed on April 20 2011 [21:18] Hello. In which "sub repository" (universe/multiverse) can I find sun-java6-jre. I'm trying to install openfire 3.7.0 from .deb downloaded from igniterealtime.org [21:19] RoyK, ty === skrewler_ is now known as skrewler [21:19] bug 634387 , i thought it was fixed.. bah [21:20] Daviey: Error: Could not parse data returned by Launchpad: timed out [21:20] RoyK Daviey - that fix works - now only 10.04.2 gives motd [21:21] guys BTW - I converted to lucid server from Centos 5.5...not going back [21:21] :) [21:22] ChmEarl: \o/ [21:23] ChmEarl: I work for an institute - our scientists need a truckload of software we earlier had to install by hand on those fscking redhat machines - now it's all (or most) available from the ubuntu repos [21:24] RoyK, only shock I had was perl modules names are polish [21:24] libmath-bigint-perl [21:24] wtf? [21:24] backwards [21:24] i18n? [21:24] oh [21:24] I see [21:24] RoyK, in RH world the name is perl-Math-BigInt [21:25] I left RH just after rh7 === NG_ is now known as ng_ [21:26] started out with slackware back in 1994 or so, then to redhat, then to debian, then ubuntu... [21:26] Any one get ldap client to work on 10.04 LTS without errors? [21:26] minkyferrari: usually by configuring it correctly :D [21:26] I am able to authenticate but auth.log says pam_unix(sshd:auth): authentication failure [21:27] I looked all through my pam modules but can't figure it out [21:27] I am able to auth though, just want to know what this failure means === j1mc is now known as j1mc_away [21:31] Hi all, I am using UEC and I have a Walrus server and I can't seem to find/mount my RAID controller and I was wondering if anyone knew if it was supported and if so how can I mount it. The controller is a Dell Perc 5i. Thanks in advance for any information you can give. [21:31] Also every 20 minutes or nslcd can't find the ldap server but then times out and reconnects [21:38] what is the appropriate path for a local apt repository? [21:38] /usr/local/mydebs is ok? [21:39] it shouldn't be accessed by anyone [21:53] bencc: What's the point then? [21:54] soren: I mean that it should be secured [21:54] I've packaged a server and I want to install it with a local repository [21:55] I can put it in /home/myuser/mydebs but it doesn't feel right === ng_ is now known as NG_ === NG_ is now known as ng_ [22:19] http://ubuntuforums.org/showthread.php?t=1767372 Can someone help ? [22:22] $ host -t AAAA podcast.ubuntu-uk.org [22:22] podcast.ubuntu-uk.org has no AAAA record [22:22] achk, wrong window, sorry [22:26] how do i make it in my sudo file such taht i dont have to enter a password for sudo [22:27] flowbee: echo "$USER ALL=(ALL) NOPASSWD:ALL" >> /etc/sudoers [22:28] well, that won't work exactly, i'd run visudo to do it [22:28] like this: [22:28] flowbee ALL=(ALL) NOPASSWD:ALL [22:28] Hmm. [22:29] the echo would work if you used sudo tee -a instead of >> [22:29] Is there a way to copy the partition tabel from one disk and map it on another? [22:29] since Im adding this disk to the array it needs to be a duplicate of the others [22:29] Slyboots: sfdisk can dump a partition table in a format that can be read back in [22:30] Slyboots: sfdisk /dev/sda | sfdisk /dev/sdb [22:30] i don't remember off the top of my head, but sfdisk(8) can help you [22:30] assumes msdos tables, of course [22:30] Mm.. I have to use GPT [22:30] then you probably need to use parted.. and i don't know that it supports something as trivial [22:31] (it didn't when we originally adopted it for systemimager, but that was yearsssss ago) [22:31] I cant figure out parted at all.. x.x [22:32] New bug: #788348 in samba (main) "package samba-common 2:3.5.8~dfsg-1ubuntu2.2 failed to install/upgrade: subprocess installed post-installation script returned error exit status 127" [Undecided,New] https://launchpad.net/bugs/788348 [22:33] Slyboots: what bit are you missing? [22:33] all of it, I need to create a bios_grub partition [22:34] But parted doesnt seem to understand what that is [22:34] bios_grub? i don't understand that either :) [22:34] * dannf has only used gpt on ia64 though [22:35] i also only use parted for partitioning - i use filesystem-specific tools for doing anything to the parittions [22:35] Its needed for grub to work in GPT partiions [22:35] Otherwsie it doesnt boot at all :P [22:35] does it just need to be a standard efi boot partition (i.e., fat)? [22:35] 1 1049kB 2097kB 1049kB bios_grub [22:35] Apparently, 1mb in size [22:37] * Slyboots fires up gparted :P === Jasonn is now known as JasonnAWAY === JasonnAWAY is now known as Jasonn [22:44] Bah..] [22:44] I may just rebuild [22:44] lol [22:45] Move the OS onto its own disk, then nuke the entire array to use it for storage only === ng_ is now known as NG_ [22:49] not sure if thats wise, but it would be a lot simpler [22:57] The person that solves my problem gets a free beer* http://ubuntuforums.org/showthread.php?t=1767372 *= free beer will not be given until 10000 years after you die [23:01] post: if you run 'sudo tail -f /var/log/daemon.log' in another terminal, you might get a clue about the problem [23:01] dannf: thanks will try it in a sec [23:09] dannf: it cant find /var/log/deamon.log :/ === med_out is now known as med === med is now known as medberry [23:10] daemon not deamon [23:11] dannf: I copy pasted what you said "sudo tail -f /var/log/daemon.log" and it gives me a error tail: cant open /var/log/daemon.log' : File not found [23:11] hm.. maybe /var/log/syslog then? [23:16] dannf: it showed "May 25 20:07:01 server1 CRON[4062]: (root) CMD (/usr/local/ispconfig/server/server.sh > /dev/null 2>> /var/log/ispconfig/cron.log) [23:16] May 25 20:07:23 server1 kernel: Kernel logging (proc) stopped. [23:16] May 25 20:07:23 server1 rsyslogd: [origin software="rsyslogd" swVersion="4.6.4" x-pid="763" x-info="http://www.rsyslog.com"] exiting on signal 15. [23:16] ^C [23:16] post: Error: "C" is not a valid command. [23:16] " [23:20] i installed squid.. it's been running great for about forever.. for whatever reason though recently it didn't start - how should i trouble shoot this? how can i make sure that the service is auto-started at boot? [23:20] andygraybeal: check /var/log/squid3/cache.log [23:20] i guess it's been running fine without issue for a year or more [23:21] check the log for any error message [23:21] they are always there [23:22] k thank you [23:22] pastebin the last lines [23:22] I know a lot of squid errors jje [23:22] and their fixes, probably I can help [23:22] k awesome, thank you, two seconds [23:23] http://pastebin.com/bCr8dj30 [23:24] i *just* started the service [23:24] you know what i mean... but i had to manually start the service... which i've never done ever... it's just run without issue for maybe a year now [23:24] post: no idea; sorry - you might check if some other dns service is already bound to that port, but hard to say for sure w/o lgos [23:24] jmedina, it's working fine after i started it.. but the thing is i had to start it :) [23:26] :) [23:27] dannf: [][][][][][][][][[] [23:28] dannf: can you recomend something better / equal to bind ? [23:29] can you try to disable apparmor and try install again? [23:29] jmedina, me? [23:30] sorry, that was for post [23:30] post: i don't necessarily think bind is the problem - it could be any number of things that might affect others as well. personally i don't have any experience w/ alternatives to bind [23:30] post: bind is the most use dns server on the net - it works [23:31] there are others that might work too, but I don't know any better ones [23:32] post: you can run this to see if something else is listening on the port: sudo netstat -pan | grep :53 [23:32] (if there's a lot of output, please paste to paste.ubuntu.com) [23:33] http://paste.ubuntu.com/612982/ [23:34] post: k - then that guess isn't your problem [23:35] post: does 'sudo grep named /var/log/*.log' output anything? [23:35] dannf: yes http://paste.ubuntu.com/612984/ [23:36] hm.. that only tells me that it tried to start, not why it bailed :( [23:37] damn'it, upgraded to natty and now my dhcp is no longer working.... [23:37] (777 permissions on named.conf are very dangerous, in case that wasn't obvious) [23:38] probably trying to fix things... [23:39] post: one last thing you can try is 'sudo named -g [23:39] ' [23:40] jmedina: oh - you were talking about this issue - yeah, might be worth disabling apparmor (and yeah, i'm sure post was just trying to fix things) [23:40] sudo named -g outputs http://paste.ubuntu.com/612986/ [23:41] post: ah - its complaining about a syntax error, see that? [23:41] 'options' redefined near 'options' [23:42] you might take a look at your config file to see what is wrong [23:42] post can you paste named.conf.options? [23:42] also the output from named-checkconf please [23:42] stupid isc dhcp server is not compiled with ldap support or what? [23:43] ldd it :) [23:48] yes, looks like so, it's not compiled with ldap support [23:48] probably you can use dpkg -s and check if depends on ldap libs or something [23:48] well, my dhcpd.conf is invalid and not working [23:48] named.conf.options http://paste.ubuntu.com/612992/ [23:49] anyway, not my problem now ;) i've got my fixed ip, maybe 60 other people may have got problems now, lol [23:50] jmedina Thx my ISP changed his DNS ips last night [23:54] Changed it to the new ones but stilll [FAIL] [23:55] mmm [23:55] what about the output from named-checkconf? [23:56] and grep ^options /etc/bind/* [23:56] jmedina I cant find named-checkconf [23:57] o_O [23:57] run it [23:58] root@server1:~# named-checkconf [23:58] /etc/bind/named.conf.options:1: 'options' redefined near 'options' [23:59] root@server1:~# grep ^options /etc/bind/* [23:59] /etc/bind/named.conf.options:options { [23:59] are thos *** lines in the options file?