[00:52] hi; having 'issues' installing grub on a server, namely, it won't install automatically and I can't see the wood for trees about how to install it manually [00:52] default guess is /sda, which fails [00:53] I made a seperate boot partition, but that seems invisible to ash, although it messes with mount points so who really knows [00:53] this is 12.4 if it matters [00:53] and yes I want to use 13.10, but it has a critical bug which prevents me [00:54] it's 12.04 btw [00:54] yes him [00:55] Hornet: so grub loads partially? [00:55] it tries to install and fails [00:55] it guesses /sda [00:56] add a /dev to that [00:56] I cannot see the boot partition in df -h so I have no idea what it did to it [00:56] well sda is in the raid anyway [00:56] check /proc/partitions [00:56] Hornet: check fstab, sometimes /boot isn't mounted [00:56] so it should be part of sde [00:56] or yeah, /proc/partitions is better [00:56] hm, hang on [00:56] this is a computer in the other room to the bloody server [00:56] and I think the command would be sudo grub-install /dev/sde [00:56] so tennis ahoy [00:57] sde has multiple partitions too, but hopefully yes [00:57] something simple like that would be refreshing [00:57] or alternatively, play musical sata ports until your disk ends up in the port for sda, instead of sde [00:57] can't, server doesn't allow that [00:57] que [00:58] doesn't really matter the device name [00:58] any bootp fanatics here? And perhaps just maybe intimate with dhcpdump? I'm seeing things I just don't understand. Like... why is my system broadcasting bootp requests? Why is it sending them to a private address? How is it that I'm actually getting responses? And why won't it stop? Here's a capture file: http://fpaste.org/83863/41241313/ [00:58] ubuntu 12.04 [01:09] jesus this is mad [01:10] looks like sde1 is the right target [01:10] but can't mount it [01:10] ash won't see grub-install as a command [01:10] nor apt-get [01:10] so .... [01:11] sheptard, disks have to stay where they are, there's only one physical location for a non-raid disk and it's in it [01:11] sde* is all the non-raid stuff [01:14] anybody? can't turn the server off until this is sorted [01:15] and will be in a worse state than bare metal if I have to start again tomorrow [01:15] afaict everything's set up other than grub [01:21] Hornet: use fdisk -l to see your partitions [01:23] apb1963: this is ash, not bash [01:23] so no fdisk [01:24] stuck at the last stage of installation [01:24] fdisk isn't builtin to the shell [01:25] apb1963: so can I use it another way? [01:26] atm it doesn't work, I don't know the details of what ash can and can't do [01:26] Hornet: I don't quite understand the question [01:26] Hornet: You typed fdisk -l and it said... ? [01:26] fdisk: not found [01:26] this isn't a live system, it's still in the installer [01:27] it has ash and presumably some things [01:27] seems grub-install isn't one of them though [01:27] so quite how I can install this system I don't know [01:27] so finish the install and then come back and install grub later [01:27] it won't boot! [01:27] why not? [01:27] no grub? [01:27] what happens? [01:28] is there an error? [01:28] I strongly expect so [01:28] if I restart and it breaks I lose hours of work [01:28] You don't know? [01:28] hmm [01:28] I can drive a car off of a cliff to see if it falls, yes [01:28] I dont know until I test it [01:28] but I have a very good idea [01:29] I'm pretty sure grub isn't required to boot [01:29] let me take a quick look... [01:29] this is a raid system [01:29] single OS, tbf [01:31] fs [01:31] in other room now [01:31] laptop didn't like that at all [01:31] and nor did the people leeching from the torrents on the external hdd [01:31] oh well [01:31] they can get their lain later [01:32] what does ls -l /dev/mapper say? [01:33] control iirc [01:33] one sec [01:33] this is mdadm raid, not bios or fakeraid [01:33] what kind of raid are you doing? bios? software? fake? [01:33] ok [01:34] crw----- 1 root root [01:34] mdstat's at 16%, to give you an idea of how long this has been set up [01:35] and this is a 6tb raid 6 array [01:35] none of the ubu server docs mention grub within the context of raid [01:35] Hornet: suggest reading all the way through this: http://askubuntu.com/questions/43036/how-do-i-install-grub-on-a-raid-system-installation [01:35] yet that's a surefire breakage [01:36] I've googled it too, yes :\ [01:36] that's a different raid setup aiui [01:37] thing is I have a partition that was meant to be used for /boot, but can't mount or select it [01:38] at the point where I may as well restart it and see what smokes [01:38] and repair if I can from a live usb stick with something that resembles a real os [01:38] otherwise start again tomorrow [01:38] fs [01:38] The gist of it that I got from a quick skim was you need to create a small ( 1 mb ) partition flagged as bios_boot to install grub on GPT [01:40] the installer did something like that automatically [01:40] but seems not to have flagged it? [01:40] basically it's halfarsed it and left me high & dry [01:41] wonder if I should start again without grub [01:41] raid even [01:41] then create manually later [01:41] more trouble than it's worth [01:41] ridiculous system [01:41] things are always a mystery until you know the answer :) [01:41] well, ash won't let me install grub [01:42] so that's basically a golden 'screw you' surely [01:42] it's now saying I'll have to boot manually passing parameters [01:43] so here comes that cliff [01:43] sigh [01:44] restarting [01:45] darthanubis: Hey. I reformatted (don't ask), and installed xubuntu server. I can mount drives, and am generally pleased with it. I swapped out firefox for midori and removed gimp, but other than that it's pretty nice [01:45] nothing there [01:45] wonderful [01:45] black screen, no dhcp allocated [01:45] dead box [01:45] told you === Guest28960 is now known as NCommander [01:45] GeekDude, glad it worked out === NCommander is now known as Guest30460 [01:47] LOL [01:47] fuck me [01:47] the stick I was planning to use to recover it, seems to be 13.10 [01:47] AKA usb keyboards die [01:47] which is all I have to use [01:47] !nomodeset [01:47] A common kernel (boot)parameter is nomodeset, which is needed for some graphic cards that otherwise boot into a black screen or show corrupted splash screen. See http://ubuntuforums.org/showthread.php?t=1613132 on how to use this parameter [01:47] jesus [01:49] it would have been on the network [01:49] I was going to ask if you had gdisk or gparted... but I guess I took too long :/ [01:49] no, it's the installer [01:49] gnome anything is unusable [01:49] so now I'll have to start all over again without raid I think [01:50] and toast all my partitions [01:50] last chance, will see if the 12.4 installer has a usable recovery anything [01:51] gdisk has nothing to do with gnome [01:51] hm, what am I thinking of then [01:51] anyway, sod all was available [01:51] so I highly doubt it [01:52] nor does gparted for that matter [01:52] in rescue mode now [01:52] need laptop charger, brb [01:52] you could have also tried cfdisk [01:53] which is curses based [01:54] did you try nomodeset? [01:54] well, in rescue mode atm [01:54] I had no opportunity to do anything [01:54] boots to nothing [01:54] !nomodeset | hornet [01:54] hornet: A common kernel (boot)parameter is nomodeset, which is needed for some graphic cards that otherwise boot into a black screen or show corrupted splash screen. See http://ubuntuforums.org/showthread.php?t=1613132 on how to use this parameter [01:54] not on network either [01:55] how could I add a kernel parameter without grub or seeing what I'm doing? [01:55] well, that is in fact a good question [01:56] the truth is out there [01:56] anyway, rescue mode can't assemble the raid [01:57] tl:dr; all fucked, abort abort abort [01:57] will restart from bare metal without raid [01:57] don't know if you know it, but there's also a #ubuntu which is more active at the moment... maybe someone there can help [01:57] it's too active [01:57] it's drowned in noise [01:57] just add to the noise... if someone sees your question... they'll try to help [01:57] thanks for the assistance though, is much appreciated [01:57] heh [01:58] sorry I didn't know the answer... lots of info out there... but it does take quite a bit of sifting. [02:05] apb1963: np, thanks all the same, and partly that's the problem [02:05] so many similar problems with inapplicable solutions [02:05] that give partial anecdote-based answers [02:06] ubuntu rescue mode is messing around atm [02:06] just as well I have nothing to lose, in theory [02:06] otherwise I'd be concerned [02:06] given my luck it'll blow the bloody fusebox nexyt [02:06] -y [02:07] yeah... unfortunately this whole OS is a bit of trial & error.... no two systems are exactly alike. Different combinations of hardware & software... lots of choices leads to a tower of babble. I've got my own frustrations :/ [02:09] I'd be happier if I were able to use a newer version [02:09] 2 year old software isn't great [02:09] 13.04 has a critical bug though, so I can't do anything with it [02:10] 12.10 would mean two upgrade hops and is too old for support anyway [02:10] so weirdly, 12.04 is the only option [02:10] Yeah, 12.04 is LTS "can't go wrong" with that solution. [02:11] You have a choice of "stability" vs. latest & greatest features with all the bugs included. It's a tradeoff. [02:11] I went with stability. Not sure I'm happy about it. [02:12] this is why I can't use debian any more [02:12] solid yes, but by the time you use anything it's basically half a decade old [02:13] only other OS I could consider is fedora [02:13] yeah, I hear ya. [02:13] admittedly a lot better than it was [02:13] but still, hardly my first port of call [02:13] Sorry but... I have a phone call to deal with.. wish you luck bud [02:13] np, thanks again [02:14] * apb1963 exits stage left [02:14] 'rescue mode' seems to be 'install everything again' mode [02:14] oh well [02:15] * Hornet ankis [02:45] okay this is mad [02:45] when I reboot it with the usb stick it, it seems to work as it should, but I can't log in [02:45] without it, it doesn't seem to do anything [02:45] usb stick is just the unetbootin installer for 12.04 [02:46] wonder if it's somehow installed grub on the usb stick [02:46] this is mad [04:07] right so, it seems that when I install ubuntu server, it latches onto the usb stick and sticks grub on there [04:07] manual reinstallation of grub fails [04:08] looks like it works, fstab looks fine too, but it doesn't [04:08] seems my device names shuffle about sometimes [08:52] rbasak, hey - what did you think about my virtual-mysql-* email? [09:03] jamespage: I've been mulling over it. [09:03] jamespage: it sounds good to me, but I'm not sure it's enough. What bothers me is the postrm. [09:04] jamespage: AIUI, as long as we have a postrm, there are two separate sources of truth about who owns what, and this seems suboptimal to me. This is limited to my understanding though. I had to refer to the policy manual. === _ruben_ is now known as _ruben [10:00] Good morning. === mld_ is now known as mld [10:46] http://ubuntuforums.org/showthread.php?t=2210294 [12:05] rbasak, https://bugs.launchpad.net/ubuntu/+source/percona-xtradb-cluster-5.5/+bug/1290331 [12:05] urgh [12:07] jamespage: any opinions on how to fix that? [12:08] rbasak, thinking about it now [12:08] jamespage: it seems like a really fundamental question to me. AFAICS, it doesn't make sense for the alternatives to share the same config file path. [12:09] Perhaps we need to hack something together for Trusty though. [12:09] rbasak, I didn't pick this up in testing because I'm using a charm which writes the entire my.cnf before installing anything [12:10] rbasak, I was just testing the virtual-* stuff and noticed this... [12:10] * jamespage sighs [12:10] Well, that's it. Charms can handle this. I don't think it makes sense for packaging to magically handle switching out of alternatives but maintaining the same DB and config directories. [12:11] That makes the assumption that they won't diverge, but being forks, clearly they will. [12:16] rbasak, I was thinking we could switch out that bit into a conf.d snippet for mysql-5.5 + others [12:16] but on a switch in/out that file would not be removed as its config right? [12:18] jamespage: I think that would be a problem, yes. Since we permit packages to be removed but not purged, and configuration files remain in that state, yet we expect a swap-out to work in that case. [12:22] rbasak, I guess the alternative is to ship with pxc disabled by default [12:22] with a README [12:22] looking at the maintainer script this should be OK [12:26] jamespage: I can't think of a better option right now. [12:36] rbasak, not sure that will work - the maintainer script want to use mysql directly if offline to manipulate the database [12:36] that fails as well... [12:42] rbasak, maybe patching pxc to ignore those values might work [12:42] * jamespage looks [12:51] rbasak, going to think about it more over lunch === dannf` is now known as dannf [13:35] hi, my ubuntu 12.04.4 lts run for several days, today i looked at it and it prints me the following command line error: "ata1: exception Emask 0x10 SAct 0x0 SErr 0x4090000 active 0xe frozen. ata1: irq_stat 0x00400040, connection status changed, ata1: SError: { PHYRdyChg10B8B DevExch } [13:35] however, after hitting "enter" the system still react and i can execute commands [13:43] what is a good energy saving program on ubuntu server 12.04.04? because i dont want ubuntu to allow to go into energy save mode, because it should always run [13:54] rbasak, I'm a dulderhead [13:55] /etc/mysql/conf.d/pxc.cnf (set query_cache_* to 0) [13:55] #OK [14:04] tomixxx3: a "program" consumes energy :-) [14:04] tomixxx3: what you want is a power saving configurators. The default powersaving modules in Ubuntu seem more or less reasonablke. [14:05] tomixxx3: as to that ATA error that is a hw error in th bus to the disk. That it worked still means that the bus to the disk was reset and that worked. [14:05] walex2: i have already installed "caffeine" [14:06] Walex2: kk [14:06] jamespage: will mysql-5.5 (for example) pick that up? What if you install percona-xtradb-cluster-5.5, then remove it and install mysql-server.5.5? [14:07] rbasak, its remaings [14:07] jamespage: so still broken but less broken? [14:08] rbasak, the pxc config is valid for mysql-5.5 - its just not the normal defaults [14:08] rbasak, yes - less broken [14:08] tomixxx3: the error may be just a random thing, or it may be a symptom of a failirg disk. [14:08] rbasak, I'm seeing this as the least worst option right now [14:08] jamespage: agreed [14:08] tomixxx3: for power management, look at packages 'powermanagement-interface' and 'pm-utils'. [14:09] tomixxx3: I have written my own script BTW here: http://sabi.co.uk/#almostSources [14:12] walex2: i need a mechanism, so that my ubuntu server 12.04.4 together with a deployed OS-GUI do NOT change to any kind of power saving mode. [14:19] jamespage, rbasak, smoser - just wanted to make sure you all had reviewed the vUDS schedule. http://summit.ubuntu.com/uds-1403/track/servercloud/ [14:19] gaughen, I just scanned through yes [14:20] gaughen, looks reaasonable. [14:21] sortiing is funy. [14:21] oh. no its not. [14:22] anyone know what badblocks output means, there are 3 numbers before errors, like so: [14:22] 17.21% done, 33:10 elapsed. (0/83943044/0 errors) [14:23] what do those 3 numbers represent? (I'm assuming regardless this drive is dead :)) [14:37] If someone could clarify... can procmail work on outgoing or only incoming? [14:41] procmail is for delivery [14:42] mardraum: OK thanks. [14:42] generally you handoff delivery to procmail [14:42] if you need it for something, that is [14:43] We want to fetch GPG keys for recipients and add them to our local database the first time they are sent something ( or we receive something from them). For incoming procmail works fine for this, outgoing seems to be another matter. [14:43] sounds like fun [14:43] which MTA? [14:43] Postfix [14:44] surprised nobody has done it before with postfix :P [14:46] Yes, it seems odd that nothing simple like procmail but for outgoing seems to be around. I looked at altermime but it seems more for rewriting/appending than for calling some arbitrary script [14:47] I wonder how stuff like ClamAV do it [14:49] genii: Using procmail to filter outgoing mail sounds like something which is possible, but non-trivial. [14:51] they likely have. === wedgwood is now known as Guest24594 [14:51] Hello, i have an question, i have 2 servers doing the same thing have the same stats the server on ubuntu 11.04 has a load of 0,05 and the ubuntu 12.04 has a load of 1.50, is there changed something in the load mesure? [14:52] genii: Writing something custom, and use milter to hook it in? [14:53] Hm. [14:55] For postfix that would be smtpd_milters/non_smtpd_milters [14:55] I have an ubuntu server, where postfix maillog ends into syslog. System is running rsyslog and ubuntu 13.10 [14:56] andol: Thanks for the tip, looking into it now [14:57] jdo_dk: See /etc/rsyslog.d/50-default.conf [14:58] andol: mail.* -/var/log/mail.log [14:59] hey guys, just installed ubuntu server 12.04 and installed desktop. does anyone ever get any kickbacks where ubuntu just throws you back to the login screen. ubuntu software center will close automatically as well after 30s of being opened [14:59] genii: what specifically would you do on outgoing? [15:05] slowcon: that looks like some hardware issue causing the X windows to crash and/or applications to crash. [15:06] jrwren: Grab the recipient address, then basically: look for keys.domain.tld or keyserver.domain.tld and if found, fetch the recipients public key from there. If not found there, fall back to public keyservers in order of pgp.mit.edu then keyserver1.pgp.com and keyserver2.pgp.com ( with ldapsearch). If a result is found, add it to the local ldap databse under recipients so that it could be looked up locally [15:06] slowcon: the hardware issues could be as simple as insufficient RAm triggering the OOM killer [15:06] slowcon: or it could be an issue with the graphics chip driver being buggy. [15:07] genii: so you want to do this action on every email which traverses your email server? [15:07] ws2k3__: probably the measure has not changed, but most likely there are subtle differences in the daemons running. [15:08] jrwren: At least on first time they are used, yes. And of course not for local-local because these are already in the database [15:11] genii: i agree with the milter suggestion. [15:11] genii: sorry for no more help than that [15:14] Walex2 well what i see is that the rest of the value's in top are the same the %us and %sy are Exacly the same [15:15] and it looks like the one server is not more the bizzy then the other the value's is just a bit higher [15:17] * genii goes back to investigating [15:39] hi, it seems i cannot use environment-variables like "export OS_USERNAME=blabla" on my ubuntu deployment (12.04.04) [15:40] because whenever i try to execute commands which should use these variables, i get a response indicating that OS_USERNAME for example is missing [15:40] tomixxx3: pastebin all the exact details. commands and output [15:42] command is "keystone token-get" vs "keystone --os-username=admin --os-password=ADMIN_PASS --os-auth-url=http://controller:35357/v2.0 token-get" [15:44] tomixxx3: that's a big jump from your earlier comment. simplify things to convince yourself you have the basics working [15:44] pmatulis: content of openrc.sh: http://pastebin.ubuntu.com/7068279/ [15:44] tomixxx3: but it sounds like a openstack/keystone problem and not a ubuntu server problem [15:45] pmatulis3: kk, ty [15:46] tomixxx3: did you source openrc.sh or are you running it? [15:46] jrwren: yes, i have executed "source openrc.sh" [15:46] tomixxx3: ok. [15:51] coreycb, if you fancy something to hack on whilst at your sprint: [15:51] https://bugs.launchpad.net/ubuntu/+source/keystone/+bug/1290423 [15:54] ws2k3__: very small differences can give slightly different load averages as they are rounded... [15:54] jamespage, thanks I'll take a look [15:55] jamespage, I need someone to review my "fix" for 1288344 [15:55] coreycb, merge proposal? [15:55] jamespage, yes, , for keystone [15:56] coreycb, keystone charm - for bug 1288344 [15:56] yes [15:56] jamespage, that's right [15:57] coreycb, ok - once you have something up ping me and I'll review [15:57] jamespage, thanks [16:01] jamespage, it's up at https://code.launchpad.net/~openstack-charmers/charms/precise/keystone/trunk/+activereviews [16:02] coreycb, looks ok [16:02] +1 [16:02] jamespage, ok wasn't sure if that was the right approach. but it's done elsewhere. thanks. [16:04] coreycb, ivoks and I redux'ed that charm completely week before last - but that work has not landed yet so your change is appropriate. [16:05] coreycb, the right way would be to have a post-start section in the upstart config that could tell when keystone was ready [16:06] jamespage, excellent that's good to know [16:21] Hm, looks like it may be possible to do outbound filtering with header_checks [16:21] * genii goes back to reading === PaulW2U is now known as G4MBY [16:33] jrwren, your cloud-init change is in -proposed now [16:33] https://lists.ubuntu.com/archives/precise-changes/2014-March/021465.html [16:34] thanks. [16:34] hallyn, congrats btw [16:35] thanks :) === martisj_ is now known as martisj === Ursinha is now known as Ursinha-afk [17:38] jamespage: do we need to remove src:samba4 from universe? It hasn't had an upload in Trusty, and src:samba seems to have overtaken it. [17:38] rbasak, good spot - yes [17:39] Looks like it has been removed from sid/jessie [17:39] I'll file a bug. [17:42] * rbasak has filed bug 1290448 [17:42] https://bugs.launchpad.net/ubuntu/+source/samba/+bug/1290448 [17:42] * rbasak wonders where the bot went. MIA? [17:43] soren: do you still run the bot (I think it was you?) [17:43] Or was that just the new bug filed bot? [17:43] * rbasak shrugs [17:47] Ubuntu 13.04 do-release-upgrade -d [17:47] Checking for a new Ubuntu release [17:47] No new release found [17:47] huh? [17:48] can't do-release-upgrade from 13.04 [17:48] what does /etc/update-manager/release-upgrades have to say about Prompt? [17:48] most likely bcause you cant go from 13.04 to 14.04 straight ... [17:49] can't get to 13.10 either [17:49] and 13.10 (your actual upgrade option) is not a devel release [17:49] Prompt=normal [17:50] so how do I so a sim[ple upgrade to 13.10 [17:50] just do-release-upgrade doesnt work ? [17:50] (no -d or any other option) [17:50] do-release-upgrade [17:50] Checking for a new Ubuntu release [17:50] No new release found [17:51] tarvid: did you check that file? [17:51] /etc/update-manager/release-upgrades [17:51] ah [17:51] yes [17:51] sorry [17:51] wierd [17:51] tarvid: what does lsb_release -a tell you? [17:52] root@horace:/etc/apt# lsb_release -a [17:52] No LSB modules are available. [17:52] Distributor ID: Ubuntu [17:52] Description: Ubuntu 13.04 [17:52] Release: 13.04 [17:52] Codename: raring === Ursinha-afk is now known as Ursinha [17:54] are you sure your system is up to date with all updates and security updates ? [17:55] that doesn't work either [17:55] that doesnt sound like an issue with update-manager then though [17:56] !pastebin | tarvid [17:56] you always need to apply all updates first before running do-release-update ... [17:56] tarvid: 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. [17:56] prettry lame for a 3 line post [17:58] 6 [18:01] my dns server screwed up === markthomas_away is now known as markthomas [19:44] am having a very unusual problem. installed ubuntu server, two nics: em1 wired ethernet, default route. wlan0 wireless. em1 is responding with its MAC for arp requests for wlan0. I have no bridge configed, nor proxy arp as far as I can tell. em1=192.168.1.4/24 and wlan0=192.168.1.3/24 pastebin config http://pastebin.com/bEiwGGrE === markthomas is now known as markthomas_away [20:42] Does Ubuntu work nicely on a Lenovo ThinkServer RD340? [20:44] ball: no idea - try [20:44] RoyK: I'll have to buy one ;-) === markthomas_away is now known as markthomas === markthomas is now known as markthomas_away === a1berto_ is now known as a1berto === markthomas_away is now known as markthomas [21:26] Does anyone know when us-east-1.ec2.archive.ubuntu.com stopped going through S3? I read it started using S3 here https://lists.ubuntu.com/archives/ubuntu-cloud-announce/2012-June/000026.html but it looks like it no longer uses S3 directly. [21:27] ianward: recently, a few weeks ago [21:27] sarnold: is there somewhere i can read more about this? [21:27] ianward: I'm not sure what the root cuase of the errors were, but the s3-based mirrors were unreliable and appeared to randomly deny pacakge downloads. [21:28] ianward: sorry, no url handy, I read it on irc [tm] :) [21:29] sarnold: ok, thanks [21:30] ugh, the irc logs are organized by date and then channel... [21:32] ianward: sorry, I'm not finding the references again. :( I believe I got the impression that the mirrors would be back "soon". [21:32] sarnold: ok, thanks for checking. good to know === markthomas is now known as markthomas_away === markthomas_away is now known as markthomas === markthomas is now known as markthomas_away [22:51] jamespage: thoughts on the networking blueprint? or discussion? https://lists.ubuntu.com/archives/ubuntu-devel-discuss/2014-March/014877.html [22:51] it would still need approval [22:52] ^open to any track leads, btw === markthomas_away is now known as markthomas