[03:54] hi, how do I make o2cb starts before /etc/fstab ? [03:54] o2cb from ocfs2-tools === diddledan_ is now known as diddledan [04:59] in which file do I need to put mount command so ocfs partition mounted at the very last during boot after all services are up, especially ocfs and o2cb? /etc/fstab does not help === markthomas is now known as markthomas|away [07:02] Good morning. === kickinz1|afk is now known as kickinz1 [07:55] hi [07:56] my home dir indie user name not open automatically help of tab press [07:56] my home dir inside user name not open automatically help of tab press [07:57] like cd /home/rozzer this is not get the user name [08:10] squisher, rbasak: if you have a repo in the right format in github.com then I can just bare clone it over and drop the current one. === kickinz1 is now known as kickinz1|afk === kickinz1|afk is now known as kickinz1 === kickinz1 is now known as kickinz1|afk === kickinz1|afk is now known as kickinz1 === kickinz1 is now known as kickinz1|afk === kickinz1|afk is now known as kickinz1 === Lcawte|Away is now known as Lcawte [09:36] hiya [09:36] can somebody please review https://bugs.launchpad.net/ubuntu/+source/openldap/+bug/1395098? it's been sitting there for a while [09:36] Launchpad bug 1395098 in openldap "Please merge openldap 2.4.40-4 (main) from Debian unstable (main)" [Undecided,Confirmed] [09:38] how do I refresh a partition size in ubuntu 14.04 when I have resized a virtual disk for a vm? without rebooting. I've tried partprobe /dev/sdc, partx -a /dev/sdc, for host in /sys/class/scsi_host/host?; do echo '- - - ' | sudo tee $host/scan; done but fdisk -l /dev/sdc still shows the old size. so far the only thing that refreshes the volume size is a reboot. [09:43] huh I think I figured it out myself, since I haven't created a partition on the disk I just fdisk /dev/sdc and executed the write command. and then re-scanned the scsi_hosts. [09:43] then it showed me the new size [09:44] if I had a partition on the disk I would have had to delete the partition in fdisk and re-create it with the same starting offset. [09:52] also TIL for host in /sys/class/scsi_disk/?\:*; do echo '1' | sudo tee $host/device/rescan; done. apparently my old way of echo '- - -' into /sys/class/scsi_host is obsolete. [10:17] smoser, can you take a look at https://code.launchpad.net/~niedbalski/ubuntu/vivid/curtin/fix-1263181/+merge/250163? [10:23] hello, can someone help me with icmp redirect problem? [10:24] i have a server, thet has two network connected to it, one is a local (A) that connects to the internet the other (B) is a subnet that in some point also connects to the internet === Lcawte is now known as Lcawte|Away [10:25] A accepts any packet, B excludes packets that connects to the 443 port [10:26] so sometimes the server will sends an icmp redirection to the B networks gateway [10:27] the iptables is configured correctly [10:27] and i allready disabled redirect accepting and sending [10:28] but the server still sends those packets [10:28] i tried this http://askubuntu.com/questions/118273/what-are-icmp-redirects-and-should-they-be-blocked [10:29] also this sysctl -w net.ipv.conf.all.send_redirects=0 (and the accept one too) [10:36] ifjGery: sysctl? [10:38] yes, it worked for like an our or so (maybe i was lucky) [10:40] ifjGery: not sure settings her ewould help -> http://blog.mattbrock.co.uk/hardening-the-security-on-ubuntu-server-14-04/ [10:45] seems like it may work, but now i get procps stop/waiting [10:46] is this normal? [10:47] sorry for my noobish questions but i usually dont work with ubuntu servers === Lcawte|Away is now known as Lcawte [10:51] ifjGery: me too lol. I just got myself in trouble with ocfs2 on ubuntu === Lcawte is now known as Lcawte|Away [12:56] hi all [13:02] can someone explain me what is the interest of upgrading manually the server kernel? for example, why upgrading to 3.19.1 ? [13:04] DzAirmaX: You want something that is in that kernel, a driver, a feature or so [13:05] DzAirmaX: if you do not need something specific to a newer kernel, dont bother. [13:19] ok thank you [13:20] but you should care that you alsways have the latest version of your kernel regardless if you use the release kernel or the backported one [13:20] (... security fixes) === Lcawte|Away is now known as Lcawte [13:51] Hi All, Having a Kernel panic during boot on a headless (cli only) Ubuntu server 10.04 .... From a grub prompt I can see the HD's and an ls (ubuntu-root)/ shows vmlinuz.old and an initrd.img (as well as an initrd.img.old...) ls (hd0,1)/ shows a WHOLE BUNCH of vmcoreinfo-2.6......-server files as well as System.map-2.6....-server files.... Anyway, system won't boot... Help! what do I do??? [13:52] System was hung but I could ssh into it... restarted, and Kernal Panic! [13:52] I suspect a borked up grub [13:52] but How can I fix / get this thing to boot in the meantime ? [13:53] from a grub CLI I can see my root fs in (ubuntu-root) [13:54] grub> ls (ubuntu-root)/ shows the root files and folders of my system [13:55] Whitor: If you can see it loading the kernel and initramfs it ain't grub. Doesn't the stack trace give some clue? [13:56] lordievader, I don't see it loading the kernel... It panics right off the bat [13:58] you could try to edit the grub like to boot vmlinuz.old and initrd.img.old.. then repair grub or install a kernel [13:59] make that grub command line [13:59] lordievader, If I boot a recovery instance... I can see a boot trace... it stops at VFS: cannnot open root device "mapper/ubuntu-root" or unknown block 0,0 [14:00] rberg_, I'm at a grub cli right now [14:00] It says please append a correct "root=" boot option... [14:01] Anyone have a grub command to do so? I'm sure I could google... but this is a production machine and I'd like to get it up asap [14:01] I didn't think it was grub... just thought I could fix it in the Grub cli [14:01] or work around it. [14:02] rberg_, how can I boot the vmlinuz.old ? [14:03] if grub2; then hit 'e' on a menu option edit the linux and initrd line then 'ctrl-x' to boot [14:03] er, how can I boot the (ubuntu-root)/vmlinuz.old ? I do see vmlinuz and initrd.img files present [14:06] rberg_, same eventual error unable to mount root fs on unknown-block(0,0) [14:07] if I were you I would boot up a rescue iso and mount everything up and have look [14:08] mmm... was hoping to point it to an older image and hopefully work around it that way... [14:09] but I've never used grub's cli before today [14:09] never had too :) [14:09] to === Lcawte is now known as Lcawte|Away [14:33] coreycb, I've dropped the pysaml2 dep for keystone for now [14:34] if it makes it for release MIR then all good - but we need to test... [14:35] jamespage, ok [14:37] damnit. vgdisplay shows me a volume group with 4TB but df says that only 2GB are there. how can that be? [14:39] df will be showing you the size of the filesystem on the LV [14:40] so I can extend to 4GB if I want, rberg_ ? [14:40] if there are "Free PE" in the VG [14:41] yee. ~2 TB [14:41] thought it runs in some RAID or so and I can't extend it ^^ [14:41] never did this. is it `vgextend`, rberg_ ? [14:42] LeMike: What exactly are you trying to do? [14:42] hmm, I dont really feel comfortable helping with that! [14:42] * koolhead17 waves [14:42] but if you have space in you VG you can use lvextend to expand the LV [14:43] then you need to also expand the filesystem ones the LV is expanded [14:43] once [14:43] LeMike: vgdisplay, pvdisplay and lvdisplay will give you a bigger picture. [14:43] LeMike: vgs, pvs, lvs for shorthand [14:44] sorry rberg_ . collizion I saw there are two 2TB hard disks in the server. so I started vgdisplay and it showed that those are listed in one volume group. I like to use this 2TB Free PE. [14:44] LeMike: What does 'vgs' display for VFree? [14:44] 1,82 collizion [14:45] LeMike: And you're wanting to use that free space to expand existing filesystems? [14:45] LeMike: I'm restating this stuff to make sure I don't help you off a cliff. :P [14:46] thanks for that collizion :) yee. there are 2GB (in the LV?) and they are full. so I want to use the remaining 1,82t [14:46] LeMike: Which LV is currently full? [14:47] the one using the VG "backup" with 1,82t free PE (saw this using `lvdisplay`) [14:48] LeMike: That LV has a name, though. I need it. :) [14:48] oh. "/dev/backups/backups" collizion [14:48] LeMike: The VG is already plenty large enough. It's the LV that's our problem. [14:50] LeMike: So you've got a line that displays like so: backups backups -wi-ao--- 2.00g [14:51] yee collizion http://i.imgur.com/P5bI4X2.png [14:53] LeMike: Okay, I think I get it. You want backups to be 3.64TB, instead of 1.82TB. [14:53] that's it collizion . [14:55] LeMike: And what filesystem is /dev/backup/backups currently? [14:57] it is "xfs (rw,noatime)" collizion [14:58] LeMike: lvextend -l +100%FREE backup/backups && xfs_growfs /dev/backup/backups [14:58] LeMike: That will expand backups to take all free space remaining in its VG, and then expand the filesystem to make that space available. [15:03] yayyy collizion . 3,7T and 1,9T available. thank you very much for that! enough to let customers place more crap in their dirs ^^ [15:04] LeMike: Glad to help! [15:04] jamespage, there's a fix in nova icehouse using oslo.utils, and we don't carry utils in trusty. it's the first use of it in icehouse afaict. [15:05] coreycb, introduces a new dependency? [15:05] I always like to learn more about maintaining servers :) that helped. thanks [15:05] jamespage, yeah [15:05] coreycb, urgh - I'm surprised that got through the stable team [15:06] jamespage, yeah me too, was thinking about opening a bug [15:06] coreycb, +1 === martinst is now known as martins-afk [15:15] My turn for a question. I've got an account on a server at a client that's showing up via 'getent passwd', but it's not in /etc/passwd. How can I determine where this account is defined, because there's obviously some remote account source in use. [15:16] jamespage, https://bugs.launchpad.net/nova/+bug/1432685 [15:16] Launchpad bug 1432685 in nova "2014.1.4 introduces new dependency for oslo.utils" [Undecided,New] [15:16] collizion: look at /etc/nsswitch.conf to see where else passwd can look [15:16] I'm going to revert the commit that introduced the dep [15:17] Pici: only files and ldap, but nothing is configured for ldap on this system. [15:18] jamespage, it's commit 4b46a86f8a2af096e399df8518f8269f825684e0 [15:20] collizion: Are you sure? [15:21] Pici: I want to say yes, but obviously I'm not sure, because the account is coming from somewhere. ldap is also defined in PAM. [15:22] Pici: Found it. [15:22] Pici: Thanks. [15:22] collizion: np [15:27] coreycb, hmm - odd [15:28] coreycb, I'd probably hack out a patch to switch that to not using oslo.utils for units.Gi [15:28] its just a literal [15:29] coreycb, but that is a mistake - oslo.utils is not in the requirements list so must be pulled in by accident [15:29] jamespage, ok I'll look into the switch [15:29] coreycb, looks failry trivial s/units.Gi/XXX/g [15:29] jamespage, k [15:40] Whitor: Are you using lvm? [15:53] lordievader, I'm not sure. I'm taking over administration of this machine... I did not do the initial install [15:54] I'm pretty sure extX was used for a fs [15:54] not sure 3 or 4 [15:54] tune2fs :) [15:54] blkid [15:54] it is virtualized.... making this a bit more difficult [15:55] niedbalski: around? I'm looking at https://code.launchpad.net/~niedbalski/uvtool/fixes-lp-1428674/+merge/251969 [15:55] Should I use a 10.04 install CD (livecd) to fix this? or should I use a 14.04? (the system running is a 10.04 [15:55] define, virtual [15:56] exactly WHAT virtualization is being used? [15:56] virtual in my case refers to VMware esxi.5.5 [15:56] so virtual has nothing to do with it, moving on [15:56] it's been virtualized in this regard for years [15:56] sounds like you just selected the wrong partition for your root [15:57] yup... just makes it harder to get into the grub menu.... gotta have fast mouse and fingers [15:57] how does that have to do with anything? [15:57] patdk-wk, I didn't select anything... this happend after a reboot [15:57] it's just as *hard* *easy* on a real physical system [15:57] did I say you selected something? [15:57] oh, I didn't mean you exactly [15:57] but someone on that machine [15:58] something made a change and it's using the wrong filesystem for root [15:58] Whitor: Before an update things worked fine? [15:58] or it's just failing initrd [15:59] lordievader, I did not perform an update... [15:59] unless one occured automatically... [16:00] but, as stated, it's been running fine for years... literally. (you can see it's v 10.04 .. [16:00] not sur ewhat running fine for years has to do with anything [16:00] running fine for years != reboot [16:01] I have seen countless systems fail to reboot [16:01] cause admins keep changing things [16:01] mmm [16:01] and since no one ever reboots, those changes are never *tested* [16:01] against if reboots work or not [16:01] Guess I'm reading wrong. [16:01] patdk-wk, we agree here [16:01] I bet a screenshot is probably needed [16:02] you keep talking about grub and kernel and stuff [16:02] not sure what exactly is failing [16:02] it sounds like it is mounting /boot instead of / though [16:06] sounds about right.... if I boot with one of the recovery options so I can see the startup messages... one of the last ones reads Please append a correct "root=" boot option ; Here are the available partitions: <=- nothing here !!! [16:06] grub> ls (ubuntu-root)/ shows the root files and folders of my system [16:07] so I'm hopefull that all is not lost yet [16:07] :) [16:07] use the normal grub item [16:07] but edit it to delete the quiet and hmm options [16:07] patdk-wk, ok trying now [16:07] whatever normally comes after quiet [16:07] can't remember, I always remove it :) [16:12] Wow... took a bit to get the shift recognized.... this system boots very fast [16:12] I don't have anything after quiet .... shall I proceed to boot? [16:12] ctrl-x [16:13] and quiet has been removed... [16:14] GRUB also has set root='(hd0,1)' Should I adjust this? [16:14] what does the whole boot line say? [16:14] ah, splash [16:14] that is the other one I kill [16:14] that *should* be fine [16:15] there are 5 lines in the edit boot options screen... you want them all? or just the linux /vmlinuz-2.6..... ? [16:15] na [16:16] just kill quiet and splash and boot it [16:17] ok, done... same message as before... Please append a correct "root=" boot option [16:18] I can't see the screen [16:18] um,, ok ? [16:18] thats the error I get ^^ [16:18] that is the only line on the screen? [16:19] thast sthe last line before the kernel panic [16:20] VFS: Cannot open Root device "mapper/ubuntu-root" or unknown-block(0,0) immedialy preceeds it [16:20] that is getting closer [16:20] I never said I cared about the last line, all the other lines tell you WHY the last line was printed [16:20] those are more important [16:21] there are 100's of messages that scroll by on startup (at least dozens) [16:22] yes, but it's a vm [16:22] it's simple to make a screen shot or video of it [16:22] that is why I asked for a screen shot [16:22] and not for you to type each one [16:22] unless you want to [16:22] ahh.... Missed the screenshot request [16:22] hold a sec, that makes more sense === markthomas|away is now known as markthomas [16:29] patdk-wk, http://i.imgur.com/IJ2RuHU.png [16:29] yeah... thats a heck of a lot easier than re-keying :) [16:30] ok, so not too useful but the md info is interesting [16:30] hmm, two harddrives on that vm [16:30] that is odd [16:30] I agree [16:31] ok, so now we just need to boot into something that works [16:31] and figure out what they did :) [16:31] did they set it up as a raid? lvm? ...? [16:31] and figure out why inird can't find it [16:31] configured a mdraid, but forgot to add it, so it's not found at reboot? [16:32] mdraid on vm is odd, but seen lots of people do it [16:32] can you get to the regular grub menu or just the recovery shell? I wonder because I am wondering if grub can find its additional stages.. [16:33] it can [16:33] but recovery shell fails [16:33] I can get to a grub menu with a list of kernel options... I can also get into edit or cli from there [16:33] and yes... recovery fails the sanme [16:33] ok cool, that eliminates a bunch of stuff [16:34] ya, it's purely a initrd issue [16:34] can you boot init=/bin/bash ? [16:34] it can't find root [16:34] how could it run bash? [16:34] ahh sure.. [16:36] VM hardware at version 9 .... if that comes into play [16:36] nope [16:39] can you boot from anythings else? like a iso and verify mapper/ubuntu-root still exists.. [16:43] rberg, Yeah... I'm setting that up now... Is there a certain version that will work better than another in this case? ...Should I use a 10.04 live CD or 12 or 14 ? [16:44] considering that the os is 10.04... [16:44] 10.04 [16:44] not sure if it really matters.. [16:44] it will, if we have to fix something [16:44] don't want mixed repo's [16:44] to figure out what is wrong, it won't matter [16:44] I would do the whole mount root / bind mount proc sys dev / chroot in and update-grub [16:44] ya, but we have to get that far [16:45] and hope the newer kernel is compat with that libc version [16:45] not stuff I want to play with :) [16:45] and hope the mess up scripts that generate grub.cfg can sort it out! [16:45] grub.cfg isn't the problem [16:45] I dont think any of that will matter for a chroot will it? [16:45] hopefully update-initrd [16:45] will fix it [16:46] heh? chroot? libc? [16:46] yes [16:46] ohh right, I gotta get that out of my head [16:47] ~/cry [16:47] dling a 10.04 image while looking for one on my internal Fileserver === JanC_ is now known as JanC [16:48] Server vs desktop iso matter? do we want a gui here? [16:48] well, server would boot faster [16:48] but no difference [16:48] this thing boots fully in like 7 seconds [16:49] It won't matter much [16:58] not a lot of seeds for 10.04 :) should be down soon though === Lcawte|Away is now known as Lcawte [17:30] Booting 10.04 - used TRy ... [17:30] At desktop .... [17:31] What can I do ?? [17:31] I see both of my HDD's ... one 255MB (boot?) partition ... and a 74GB data partition [17:32] patdk-wk, ^^ [17:32] also lordievader [17:33] get to command line [17:33] and run blkid [17:36] doing ... Blkid returns nothing [17:36] blkid [17:36] heh? [17:36] yeah [17:36] are you root? [17:36] let me sudo it... I'm in the live cd [17:37] yeah... [17:37] now we've got stuff [17:37] sda = ext2 sdb = LVM2_member [17:37] both have unique UUID's [17:38] and would be a pita to re-key :) [17:42] /dev/loop0: Type=Squashfs then /dev/sda1: UUID="long unique ID" TYPE="ext2" then /dev/sdb1: UUID="another long unique ID" Type="LVM2_member" [17:42] then returned to a shell prompt [17:43] So, what can I do with this blkid stuff ? [17:44] what about pvs and vgs? [17:44] well, it's saying your root is on sdb or sdb1 [17:44] strange it shows both of them [17:44] I don't see those other things [17:44] did you type them in? [17:45] oh, let me try [17:45] are those commands? they arn't recognized if they are [17:46] sudo blkid gives me what I keyed in ^^ [17:46] so you need to install the lvm package [17:46] I'm running the live image [17:46] is that ok here? [17:46] apt-get install lvm2 [17:46] yes [17:46] sure... one sec [17:47] ok... let me post a screen cap. I've got output on those cmds now [17:50] http://i.imgur.com/M4bBraz.png [17:50] patdk-wk, ^ [18:02] Whitor: lvs, then [18:02] ok.. so any suggestion on what I can do? [18:04] hmm [18:04] that makes sense [18:04] now we need to set it all up I guess [18:04] fun fun fun [18:04] mount /dev/mapper/ubuntu-root /mnt [18:04] mount /dev/sda1 /mnt/boot [18:07] ok... the first one didn't take ... Mount: Special device /dev/mapper/ubuntu-root does not exist [18:08] Am doing these with sudo btw [18:08] Whitor: what does 'lvs' have to say? [18:09] Under LV I've got Root and Swap_1 on separeate lines [18:09] mount /dev/ubuntu/Root /mnt [18:10] or /dev/mapper/ubuntu-Root if you like - it's the same thing [18:10] oh? it's uppercase? [18:11] no ... sry [18:11] lowercase [18:11] Whitor: ls /dev/ubuntu [18:11] In /dev/mapper I only have one dir... control [18:11] someone offlined it? [18:11] shrug [18:11] lvs -a y ubuntu/root [18:12] lvchange -a y [18:12] heh [18:12] patdk-wk: I've seen LVs come up offline with live boot of ubuntu [18:13] sudo lvchange -a y returns: Please give logical volume paths [18:14] Whitor: sudo lvchange -ay ubuntu [18:14] lvchange -a y ubuntu/root [18:14] that took! [18:15] unmount /dev/sda1 [18:15] mount /dev/mapper/ubuntu-root /mnt [18:15] mount /dev/sda1 /mnt/boot [18:15] umount ? [18:16] well, you said the second one worked [18:16] or well, only the first one errored [18:16] though both should have [18:17] got it... doing ... [18:17] ok... all of them took [18:17] Making progress?!! [18:17] good lord, I'm going to owe each of you a pizza or something [18:18] :) [18:18] mount -o bind /dev /mnt/dev [18:18] mount -o bind /sys /mnt/sys [18:18] mount -o bind /proc /mnt/proc [18:18] all three of those commans worked... I just change unmount to umount ... [18:18] mount -o bind /dev/pts /mnt/dev/pts [18:18] doing the mount -o now [18:19] oh ya, my typoe :) [18:19] think that is all we need [18:19] chroot /mnt [18:20] patdk-wk: you could probably have just chrooted after the initial mounts and then just ran a mount -a [18:21] dunno :) [18:21] on the last one ... pts ... I got Mount: special device /pts does not exist [18:21] probably won't matter [18:21] ok... so chroot /mnt now ? [18:21] Whitor: not /pts, /mnt/dev/pts [18:21] Is /dev (r)bind mounted? [18:22] mount -o bind /dev/pts /mnt/dev/pts [18:22] oh shiznit ... my turn for the typo [18:22] royk, no, wouldn't work, might if you atleast do /dev though [18:22] patdk-wk, yeah when I typed it properly ... it took :) [18:22] patdk-wk: worked last I tried [18:22] heh [18:23] might be fixed up, it didn't used to [18:23] but then, that was probably as far back as 7.04 [18:23] some time ago ;) [18:23] if the old way isn't broken? why bother learning? :) [18:23] ok.... so just did sudo chroot /mnt ..... It took... What next? [18:24] hmm [18:24] cat /etc/fstab [18:24] maybe also cat /etc/defaults/grub, too [18:24] !pastebin [18:24] my prompt changed to root@ubuntu:/# [18:24] For posting multi-line texts into the channel, please use http://paste.ubuntu.com | To post !screenshots use http://imgur.com/ !pastebinit to paste directly from command line | Make sure you give us the URL for your paste - see also the channel topic. [18:25] Whitor: that's correct [18:25] likeky, I'm thinking, update-initramfs -u -k all, and update-grub, will fix it up [18:25] but jus twant to check a few things first [18:26] oh, and df -h [18:26] hopefully not out of diskspace on /boot [18:26] btw, it looks like /dev/sda1 is FULL! Doing a pastebin of my current terminal .... hold a sec [18:26] oh, you are :) [18:28] I bet that was your problem all along.. the initrd was truncated [18:28] http://i.imgur.com/h8Vyz1O.jpg [18:29] Thats me from my other box .... cant cut and paste into a pastebin as I am in a vm [18:29] ok, so looks like we are probably just fine [18:29] but you'll see the text in question there [18:30] can you do a ls /boot/ [18:32] here is /default/grub :http://imgur.com/3ejn2Bq [18:32] ls /boot/ gives a whole bunch of initrd and vmlinuz files [18:33] yep [18:33] I suppose this is good ? [18:33] well, I want to know exactly what ones [18:33] so I can tell you want ones to *delete* :) [18:35] http://imgur.com/D3hAll6 [18:36] Thats most of them... I can take two screen caps if we need to see them all [18:36] apt-get remove linux.*2.6.32-[45].* [18:37] 19 kernels installed :D [18:38] that is what auto-updates are for :) [18:38]