[00:20] <zartoosh__> hi I have done someting bad that my system during boot hangs, I like to boot and run one service at a time, what is the boot option for that so I know where it hangs? thanks
[00:21] <riz0n> Hello guys, I am running 12.04 LTS on my Linux web server, and want to upgrade to 14.04 using the do-release-upgrade but it says that there are no new releases of Ubuntu available. Anyone know when this release will be available through online upgrade?
[00:21] <sarnold> riz0n: I thought that switch was flipped earlier today?
[00:21] <riz0n> Lemme check :)
[00:22] <kklimonda> it should be, my 12.04 server is definitely telling me about it
[00:22] <riz0n> Checked it at 9 this morning, no go, and now it's there :) WOO HOO :)
[00:25] <riz0n> Wow must be a lot of people getting Ubuntu today. can only get 50kB/s... anyways we will wait and get :)
[01:10] <Patrickdk> riz0n, heh, it's always fast for me, but then I have a local mirror
[02:22] <photon> Hi. Two hours ago I unmounted an ext4 partition. I tried to remount it half an hour ago, got an error (don't remember the exact wording, but IIRC something about bad descriptors was in it). dmesg showed "group descriptors corrupted" for that partition. I ran fsck on it. It fixed tens of thousands of "Group descriptor ... checksum is invalid." errors, tens of thousands of "Free blocks count wrong for group" errors, and just as many "Inode bitmap
[02:22] <photon>  differences" errors, followed by another bunch of "free inode count wrong" errors. After an hour of fixing, I got no more errors and remounted the partition. At first glance, it seems all my files are still there and intact. What could have caused this? Could it be that the HDD is failing? SMART shows no such indication and no errors. Should I buy a new HDD? Should I reformat? Or can I safely continue to use this partition? Thank you!
[02:54] <^Mike> I currently have 2 x 2TB and 2 x 1TB in a 4-bay enclosure. 1 x 2TB is in use, with LVM + LUKS (1PV, 1VG, 1LV, 100% allocated). I'd like to move towards raid5 or 6. Is it possible to upgrade this single disk to raid5? Is it possible to upgrade LVM from raid5 to 6? To do any of this, do I need to replace the 1TB drives with 2TB?
[02:56] <^Mike> Also, if you have a pointer to complete and up-to-date documentation on LVM, I'd love you to share it with me. I understand RAID is a late addition to LVM, and it doesn't appear in many of the guides I'm finding.
[03:02] <^Mike> "For RAID 5 and RAID 6, you must use md [and not lvm] to implement software RAID." - textbook I won't be buying
[03:10] <riz0n> OK So I have upgraded my server to new LTS. And now the PHP on my web servers fail to run. What do I need to change in my conf files (Apache, I imagine) to get php to run once again?
[03:32] <^Mike> riz0n: what does "fail to run" mean? What do your logs show?
[03:39] <riz0n> My logs show nothing. When the page loads, the php code is simply mixed in with the HTML code.
[03:41] <RoyK> riz0n: probably the php module not loaded
[03:42] <RoyK> riz0n: check /etc/apache2/mods-enabled
[05:49] <chriys> hey guys does someone knows how to setup rpm repository ?
[05:52] <Abhijit> chriys, there is something called as mrepo. check if it works in ubuntu
[05:53] <chriys> Abhijit: in Ubuntu there's Alien that install them but I can't manage to find where the package is install
[05:54] <chriys> and if I run apt-get install package-name it says it's already installed
[05:54] <Abhijit> chriys, you dont need to install rpm packages in order to create its repo?
[05:54] <jpds> chriys: Wait, you want to install tpms, or create a repo of them?
[05:54] <Abhijit> chriys, whats your actual aim? you want to host rpm repo or you want to install rpm in ubuntu?
[05:56] <chriys> I think I'm lost I used this command to install a repo alien -i http://installrepo.kaltura.org/releases/kaltura-release.noarch.rpm to set up a repo
[05:56] <chriys> but after that if I run apt-get install package-name it says that the package it's already installed
[05:56] <chriys> but I can't manage to find where the package is installed
[05:57] <jpds> Does alien install packages?
[05:57] <jpds> I thought it just converted them.
[06:00] <chriys> jpds: in any case I can't find where the package is
[07:58] <Sling> trying to make my first upstart job using start-stop-daemon, but all it does so far is hang, details http://paste2.org/y7I47APt
[07:58] <Sling> then when I ctrl-C, and start it again, it tells me the job is already started
[07:58] <Sling> while it really didn't start anything
[08:09] <tortib> Hi, is anyone awake?  I'm trying to hide processes from other users using the hidepid=2 option for /proc ; I have mounted proc with the following and it appears to be on but my users still can see all the procs...proc on /proc type proc (rw,hidepid=2)
[08:10] <memoryleak> Hi. I have a folder /shared which is a mount point of SAN storage. I would like to test write and read speeds, is there a recommended way to do that?
[08:18] <klep> anyone have experience with ubuntu provisioning nodes with MAAS? I run into this issue when I'm doing bulk provisions "network autoconfiguration failed your network is probably not using the DHCP protocol" ... this only happens when I do multiple nodes ...
[08:18] <klep> any ideas?
[08:54] <h1r3> hello... my server reboots randomly without apparent reason... i looked at the logs and the only thing i found that looked kind of suspicios was the following: "kernel: [   50.309356] init: plymouth-upstart-bridge main process ended, respawning". but still i don't even know if it is related...
[08:55] <Sling> looks like this channel is just questions and no answers =)
[08:56] <cfhowlett> Sling, feel free to contribute your knowledge by answering
[08:57] <Sling> cfhowlett: I actually entered with a question too ;)
[08:57] <Sling> but had to go afk
[09:02] <h1r3> i already checked if it's a panic issue, which is not... i am gonna check if it has to do wit the temperature... i am also thinking that it has to do with owncloud, since the reboots happen apparently when it is running... i would appreciate some advice...
[10:35] <hxm> how to give permissions to scripts to access apache logs?
[10:36] <hxm> is it a good idea to change the owner of those files?
[10:42] <peetaur2> hxm: use the principle of least privilege. Give it read only if it must have it, but not write unless it also must have that (and it very likely does not ever need write access)
[10:42] <peetaur2> owner can do all... rm, chmod, etc. so it is too much
[10:51] <RoyK> hxm: if the script runs as a user member of the adm group, it should work well
[11:21] <jamespage> jodh, around?
[11:23] <DenBeiren> i have two new disks in a  running system that i would like to put in raid1 and use as /backup
[11:24] <DenBeiren> i'm guessing i overlooked a step
[11:24] <DenBeiren> first i did a fdisk, then make2fs and then mdadm
[11:24] <DenBeiren> is this correct?
[11:25] <peetaur2> mkfs last
[11:25] <peetaur2> fdisk makes a partition. raid makes partitions into raid. mkfs makes a device (your raid) into a file system.
[11:26] <peetaur2> and keep in mind that the best backup is elsewhere, in case of fire, and also without granting access to delete the files in case the backed up system goes rogue
[11:27] <RoyK> DenBeiren: no reason to use partitions - just mdadm --create --level=1 --raid-devices=2 /dev/md(something) /dev/sd(something) /dev/sd(something)
[11:28] <DenBeiren> i used assemble instead of create
[11:28] <RoyK> assemble is for assembling existing raids
[11:28] <DenBeiren>  sudo mdadm --create --verbose /dev/md0 --level=1 /dev/sda /dev/sdb
[11:28] <DenBeiren> mdadm: no raid-devices specified.
[11:29] <RoyK> mdadm --create --verbose --level=1 /dev/md0 /dev/sd[ab]
[11:29] <RoyK> btw, where's your root partition?
[11:29] <RoyK> that's usually on sda
[11:29] <RoyK> so don't do that yet
[11:29] <DenBeiren> on a hardware raid 5 part in the server
[11:29] <RoyK> what dev?
[11:30] <DenBeiren> sdc
[11:30] <RoyK> ok
[11:30] <RoyK> go on, then
[11:30] <DenBeiren> same error
[11:31] <DenBeiren> no raid devices specified
[11:31] <RoyK> ah
[11:31] <RoyK> sorry
[11:31] <RoyK> add --raid-devices=2
[11:31] <RoyK> (as I said initially)
[11:32] <DenBeiren> there we go :-)
[11:32] <pmatulis> morning
[11:32] <RoyK> DenBeiren: then, again, if I were you, I'd use LVM on top of that
[11:33] <DenBeiren> the R5 uses lvm
[11:33] <RoyK> DenBeiren: more flexible in case you need to change things later - just vgcreate yourvg /dev/md0
[11:33] <RoyK> good
[11:33] <RoyK> then lvcreate ...
[11:34] <RoyK> better allocate what you need for the first lv and then grow it later if you need more space - you never know when you want another filesystem for something
[11:34] <RoyK> or - if you want to extend the existing VG, well, it should be safe
[11:34] <DenBeiren> it's only used as internal backup
[11:35] <DenBeiren> so no need to shrink/grow later on
[11:35] <DenBeiren> the R5 has plenty of room,.. +/- 25 times what they need :-)
[11:37] <DenBeiren> so after creating,.. assembling?
[11:38] <RoyK> DenBeiren: no, after it's created, you just have to put lvm+filesystem or just a filesystem on it
[11:38] <DenBeiren> eeerrm,.. just caught my eye,..
[11:38] <DenBeiren> the raid 5 is called md0
[11:38] <DenBeiren> and i just created a second md0 ?
[11:39] <RoyK> pastebin /proc/mdstat
[11:39] <RoyK> hwraid doesn't have an mdX
[11:39] <RoyK> it just shows up in linux as a scsi device
[11:40] <DenBeiren> http://pastie.org/9465660
[11:42] <RoyK> DenBeiren: that looks like a 2x2TB mirror being synced
[11:43] <RoyK> DenBeiren: now, run
[11:43] <RoyK> mdadm --detail --scan >> /etc/mdadm/mdadm.conf
[11:43] <RoyK> and
[11:43] <RoyK> update-initramfs -u
[11:44] <hadifarnoud> I remember there was a way of adding a new virtualhost with one command. You know what that was?
[11:45] <RoyK> apache virtualhost?
[11:45] <zartoosh__> hi I am using ubuntu 14.04. I am looking for a tool or mechanism to force kernel to re-read UUID information from storage devices? thx
[11:46] <DenBeiren> RoyK: http://pastie.org/9465748
[11:49] <RoyK> DenBeiren: sudo -i first
[11:49] <RoyK> DenBeiren: with that syntax, mdadm will be run as root, but the redirect will be run as the current user, so it won't work
[11:51] <hadifarnoud> is there any script for adding virtualhost
[11:51] <hadifarnoud> ?
[11:51] <RoyK> hadifarnoud: apache?
[11:51] <DenBeiren> now i get unknown keyword
[11:51] <hadifarnoud> yes RoyK
[11:51] <RoyK> hadifarnoud: a2ensite
[11:51] <hadifarnoud> I found a few. didn't work
[11:52] <RoyK> hadifarnoud: but you need to configure it first in /etc/apache2/sites-available/
[11:52] <hadifarnoud> RoyK: that does everything for me?
[11:53] <RoyK> just read up on configuring apache - it's not that hard - take a look at /etc/apache2/sites-available/default
[11:53] <RoyK> DenBeiren: sudo -i
[11:53] <RoyK> DenBeiren: then run mdadm ...
[11:53] <hadifarnoud> RoyK: I understand I need to copy that in same folder. like /domain.com
[11:54] <RoyK> hadifarnoud: cd /etc/apache2/sites-available/ ; cp default yoursite.com
[11:54] <RoyK> edit yoursite.com with your favourite editor
[11:54] <DenBeiren> http://pastie.org/9465853
[11:54] <RoyK> enable it with the command I gave you (which really just symlinks that file to /etc/apache2/sites-enabled/, which you can do manually), and run apache2ctl graceful
[11:55] <hadifarnoud> ok thanks
[11:55] <RoyK> DenBeiren: huh?
[11:55] <RoyK> root@u1404:~# mdadm --detail --scan
[11:55] <RoyK> ARRAY /dev/md0 metadata=1.2 spares=1 name=u1404:0 UUID=86c59c02:20b550e5:e2a29b8c:73717304
[11:56] <DenBeiren> ARRAY /dev/md0 metadata=1.2 name=sax:0 UUID=7818e495:6b4d8a56:9336b751:4547158c
[11:58] <DenBeiren> can i run update-initramfs -u before the array is completely built?
[11:59] <RoyK> DenBeiren: add that line to /etc/mdadm/mdadm.conf
[11:59] <RoyK> DenBeiren: just edit the file
[12:00] <DenBeiren> and what does it do exactly?
[12:00] <RoyK> DenBeiren: update-initramfs -u isn't related to the array status - it just updates the initrd/initramfs so that mdadm.conf is copied into it and md0 is called md0 at restart
[12:00] <RoyK> without it, it'll probably be called md127 or something
[12:01] <RoyK> another good reason for using lvm - if mdX changes name, lvm will still find its way and filesystems will mount correctly - if mdX changes its name and you have /dev/md0 in your /etc/fstab, it won't mount it and probably hang waiting for input during bootup
[12:02] <DenBeiren> uhu
[12:02] <DenBeiren> so http://pastie.org/9465954 this looks ok to you? no md0 overwrites or something?
[12:03] <RoyK> I don't quite see why you should need partition tables on sd[ab], but then, it won't hurt - they'll probably be overwritten - but it looks good
[12:04] <RoyK> just vgcreate mybackupstorageorsomething /dev/md0
[12:04] <DenBeiren> ok
[12:04] <RoyK> then something like
[12:04] <DenBeiren> it's the first time i create raid array "not" using the livedisc installer
[12:04] <RoyK> lvcreate -l100%FREE -n mybackupstorageorsomething mybackupstorageorsomething
[12:06] <RoyK> the -n gives the name of the LV, the last argument is the VG name
[12:06] <RoyK> man lvcreate
[12:09] <RoyK> DenBeiren: there's always a first time ;)
[12:10] <DenBeiren> thanks for helping out in any case :-)
[12:10] <RoyK> :)
[12:11] <DenBeiren> resync 2TB takes ages :-)
[12:11] <RoyK> I know...
[12:14] <smoser> hallyn, "enabled" ?
[12:14] <smoser> in the kernel ?
[12:17] <DenBeiren> RoyK: any exp. with zentyal ?
[12:17] <RoyK> nope
[12:18] <DenBeiren> it's a gui to put over ubuntu server
[12:18] <DenBeiren> but i'm having problems with samba
[12:19] <DenBeiren> https://forum.zentyal.org/index.php/topic,22904.0.html
[12:19] <RoyK> I know - buy why?
[12:19] <RoyK> learn the commandline!
[12:19] <RoyK> you'll end up doing that anyway
[12:19] <DenBeiren> it's the company that uses that server that wants it :s
[12:19] <DenBeiren> they want to add users and shares themselves
[12:20] <DenBeiren> and they don't know any commands nor have they heard of terminal :-)
[12:21] <RoyK> DenBeiren: try #zentyal
[12:21] <DenBeiren> channel is dead,.. :s
[12:21] <DenBeiren> been there for a couple of days,.. hardly anyone on there
[12:22] <RoyK> DenBeiren: sorry - no idea about zentyal
[12:23] <DenBeiren> ok,.. thx anyway ;-)
[12:24] <rbasak> rcj or utlemming: is there a Launchpad project to file bugs in the cloud images against?
[12:24] <rbasak> eg. bug 1354839
[12:25] <jodh> jamespage: hi
[12:25] <jamespage> hey jodh
[12:26] <jamespage> jodh, I think I know the answer - its not possible to provide a custom 'restart' script in an upstart configuration right?
[12:27] <jodh> jamespage: there is no built-in support for that, but you could always pass in a variable to modify the behaviour maybe if this is a manual restart?
[12:27] <jamespage> jodh, ok - that's what I thought
[12:27] <jamespage> jodh, and a variable was one way around that
[12:31] <rcj> rbasak, got the bug, we don't have a public project which is problematic for Vagrant images
[12:34] <rbasak> rcj: maybe create a Launchpad project to track cloud image bugs for all cloud images?
[12:35] <rbasak> rcj: for example I filed bug 1355343 yesterday, and in the first instance I wanted to file against cloud images generally.
[12:35] <rbasak> (though this one only affects cloud images - any fix would be solely in cloud-init I think)
[12:38] <rcj> rbasak, and in the case of that cloud-init bug that's not a problem with how we create the image, but an issue with that package.  So it's the right way to go.  But back to the project, I think we need something but it'll be hard for users to know when to file against it.  I'd like to work this out with utlemming.
[12:44] <rbasak> rcj: sure, spend some time working it out. I agree it would be hard for users to know, but bug triagers and other Ubuntu devs will be able to help. At least you'll have all the bugs in one place then.
[12:46] <theToastIsDone> how do i create an A record on a newly created Ubuntu server?
[12:47] <ikonia> theToastIsDone: you create it on the dns server
[12:47] <ikonia> not your ubuntu server
[12:47] <theToastIsDone> ok th
[12:47] <theToastIsDone> thx
[12:49] <theToastIsDone> ok, this is weird.. if my domain registrar has private nameservers registered (ns1 and ns2.mydomain.com), and mydomain.com is pointed to those 2 nameservers, where would I add the A record, at the registrar someplace?
[12:49] <ikonia> theToastIsDone: you talk to your registrar to either a.) update their dns servers for you b.) tell them to point your domain at dns servers you maintain
[12:49] <ikonia> if you're asking this question I suspect option 1 would be wise
[12:53] <elliotd123> I've upgraded to 14.04.1 from 12.04.4 and now I'm getting "mount: unknown filesystem type 'ext2' on my /boot mount point. Any fix for this?
[12:55] <ikonia> elliotd123: what poit are you getting that error ?
[12:55] <ikonia> elliotd123: eg: after grub during the initial boot, or is grub erroring
[12:55] <elliotd123> it drops into a root shell, and if I try to mount -a I get it
[12:55] <ikonia> is it ext2 ?
[12:55] <elliotd123> after grub
[12:55] <ikonia> as in the file system
[12:56] <rbasak> jamespage: do we have an accepted way to flag bugs that probably need looking at by us later? Eg. bug 1355253. I'm wondering whether to just assign ~ubuntu-server or something, so I can find these myself later.
[12:56] <rbasak> But then again I don't want to imply that others shouldn't work on it.
[12:56] <elliotd123> I think it's ext2
[12:56] <rbasak> Or imply to the submitter that it will definitely get looked at.
[12:56] <ikonia> elliotd123: you "think" ?
[12:59] <elliotd123>  yeah how do I tell? It's our clients VM so I can't rule out that they changed something...ugh
[13:00] <ikonia> elliotd123: manually mount it
[13:00] <ikonia> did you not check any of this before upgrading a clients vm ???
[13:00] <elliotd123> they upgraded it
[13:00] <ikonia> elliotd123: you said "I've upgraded"
[13:00] <elliotd123> sorry
[13:01] <elliotd123> I figured it out. thanks.
[13:01] <ikonia> cool
[13:01] <elliotd123> looks like their upgrade never completed.
[13:04] <elliotd123> so: enabling the network and "apt-get update && apt-get -f install" looks like it worked.
[13:17] <zartoosh__> HI for uefi boot, I need to create startup.nsh and place fs1:\EFI\bootid\grubx64.efi   Does anyone know how to create it and place it where? thx
[13:39] <jamespage> rbasak, I'd target it either to a milestone for a release or raise a series task for utopic
[13:44] <theToastIsDone> is there a webmin channel?
[13:44] <theToastIsDone> does anyone know?
[13:45] <cfhowlett_> !webmin|theToastIsDone,
[13:45] <theToastIsDone> hmm... crap..
[13:45] <theToastIsDone> how do i uninstall it then??
[13:45] <cfhowlett_> theToastIsDone, sudo apt-get purge webmin
[13:46] <theToastIsDone> nice thx
[13:46] <cfhowlett_> theToastIsDone, that only works if you used apt to install
[13:46] <ikonia> how is it getting installed if it's not in the repo
[13:46] <cfhowlett_> ikonia, dpkg ?
[13:46] <cfhowlett_> source?
[13:46] <ikonia> I don't know....
[13:47] <theToastIsDone> you had to add the repo
[13:47] <theToastIsDone> which is something I'm fixing now
[13:49] <theToastIsDone> anyone know of a good webmin-like piece of software?
[13:49] <theToastIsDone> for ubuntu of course
[13:51] <ikonia> there really isn't a "good" one
[13:51] <ikonia> there are various bad ones
[13:51] <theToastIsDone> lol ok.. what's the best 'bad one' that is open source?
[13:52] <pmatulis> theToastIsDone: zentyal has a bunch of modules (apt-cache search zentyal-) but zentyal is mostly meant to be used as an install/ISO
[13:52] <pmatulis> (so i'm not sure about using them independently)
[13:53] <theToastIsDone> ok thx guys
[13:54] <cfhowlett_> theToastIsDone, might ask ##linux as well
[13:56] <DenBeiren> theToastIsDone: check zentyal
[13:59] <hallyn> smoser: yes, i fired up an ec2 instance, couldn't create a bridge, and got -enomem when i tried to modprobe bridge
[13:59] <theToastIsDone> zentyal looks nice
[14:02] <smoser> hallyn, more info? instance type ?
[14:02] <hallyn> ami-5ebf6a36
[14:03] <hallyn> and ami-a427efcc
[14:03] <hallyn> i used the default mt, presumably t1.micro
[14:04] <zartoosh__> HI for uefi boot, I need to create startup.nsh and place some boot order.   Does anyone know how to create it and place it where? thx
[14:05] <smoser> hallyn, t1.micro isn't default. probaly m1.small is. but that woudl depend on how you launched it.
[14:06] <smoser> but anyway.... it just doesn't seem likely. can you show me what you did?  the kernel is the same as we use everywhere else.
[14:06] <smoser> ie, the one on your laptop probably.
[14:11] <hallyn> I did ec2-run-instances -k sergeh-keypair ami-5ebf6a36
[14:11]  * hallyn does it again
[14:12] <hallyn> wht's your lp-id (and why is that not in the directory)
[14:12] <hallyn> i see, it's smoser
[14:57] <smoser> hallyn, i can verify that 'apt-get install lxc' didn't work on the alpha2 image for utopic that you posted.
[14:58] <hallyn> smoser: this time i can't even log into the image
[14:58] <hallyn> smoser: is my ami supposed to be a valid one (came from ubuntu-cloudimg-query)
[14:59] <smoser> http://paste.openstack.org/show/93906
[14:59] <smoser> your ocnsole output might show similar things
[14:59] <smoser> stack traces
[15:00] <theToastIsDone> here's a cron question... If I have a key setup in a get statement in cron, nobody can see that from the public side of things, correct? for example:  wget example.com/?some_key=1234
[15:02] <peetaur2> theToastIsDone: you can see that with ps -ef
[15:03] <smoser> hallyn, ^
[15:04] <smoser> theToastIsDone, you can prevent 'ps' from showing it just by putting the 'wget' into a file that is executable and only readable by the owner of the file.
[15:05] <smoser> oh. wait.
[15:05] <peetaur2> no you can't ....
[15:05] <smoser> no. never mind. the wget would still be reaable. :)
[15:05] <peetaur2> but you can store it in a file, then find some argument to wget that reads the file directly, without the content ever going on a command line
[15:08] <theToastIsDone> ok, i appreciate it everyone
[15:09] <hallyn> smoser: yeah, http://paste.ubuntu.com/
[15:10] <stetho> Hi. Does anyone know if there's something "special" you need to do with the DHCP option ntp-servers? My company doesn't allow NTP requests in or out and has its own NTP servers. My DHCP server has option ntp-servers pointing to the internal NTP servers but when I install Ubuntu it waits a long time at the "getting the time from an NTP server" message and, according to the logs, fails. After install is completed if I manually set the servers in
[15:10] <stetho>  ntp.conf everything works fine. So it would seem that the Ubuntu installer is ignoring the option ntp-servers setting or I'm missing another step that I need to take.
[15:13] <rbasak> kickinz1: FYI, I filed bug 1355890. If it gets approved, we'll have a little more time for bcache.
[15:13] <rbasak> kickinz1: no more action needed on the bug for now. The release team will look at it.
[15:14] <smoser> hallyn, https://bugs.launchpad.net/ubuntu/+source/linux/+bug/1355891
[15:14] <rbasak> stetho: which installer are you using?
[15:14] <rbasak> stetho: there is a specific d-i feature that should support DHCP supply of NTP servers.
[15:15] <smoser> hallyn, did you think you saw this on trusty ?
[15:16] <hallyn> smoser: I don't remember what happend on trusty, lemme restart that
[15:16] <smoser> hallyn, i'll check. i just launched one.
[15:16] <hallyn> smoser: my second start of htat utopic ami again ended up refusing to let me login
[15:17] <rbasak> stetho: IIRC, dannf wrote that support. I can't find it right now - perhaps dannf can remember where it was or what the bug was?
[15:17] <hallyn> I assume it's just a matter of when the kernel crashes
[15:17] <smoser> hallyn, yeah, clearly kernel not working could cause failure of ssh
[15:18] <stetho> rbasak: This is just using the ISO. This is more about "what does option ntp-servers" actually do rather than finding a way round it. I'm trying to understand if it's a setting for another OS or if there's a reason Ubuntu ignores it.
[15:19] <rbasak> stetho: I know that there is a DHCP option for which support was specifically added for your use case.
[15:19] <rbasak> I'm pretty sure it's ntp-servers.
[15:19] <rbasak> stetho: I can't find the specifics right now though, which would help actually fix the issue for you.
[15:21] <smoser> hallyn, i cannot reproduce on trusty
[15:21] <smoser> at least not in the sample set of 1
[15:21] <kickinz1> rbasak: thanks, Right now, I'm benchmarking bcache with writeback enabled. I'll keep you informed.
[15:22] <hallyn> smoser: ok, good to know it's not my setup tha tfailed me with the utopic one
[15:22] <hallyn> (i'd just had to switch my credentials around on my new laptop, wasn't quite sure)
[15:22] <stetho> rbasak: I don't seem to be able to find anything either which is why I asked in here :-)
[15:27] <mic_> anyone using nginx? I have a few short questions.
[15:29] <hallyn> jamespage: around?
[15:29] <jamespage> hallyn, yes
[15:30] <hallyn> jamespage: ipxe in utopic fails to build.  the reason is syslinux does not ship /usr/lib/syslinux/isolinux.bin.  That's easily fixed by adding the install line for that to debian/syslinux.install.  Right now it only ends up in the syslinux-udeb package.
[15:30] <hallyn> jamespage: so I intend to push that to utopic,
[15:30] <jamespage> hallyn, ok
[15:30] <hallyn> just pinging you first bc you may have an idea of why i'm being sily
[15:30] <hallyn> silly
[15:41] <jamespage> hallyn, just so I understand the context, when did this stop working?
[15:41] <jamespage> does ipxe fail in debian as well for the same reason?
[15:41] <hallyn> jamespage: it does, i tested it in jessie last night
[15:42] <hallyn> and tested that with that fix, ipxe did build
[15:44] <punkgeek> how to install java for run java script on apache?
[15:48] <jamespage> hallyn, and syslinux only ships that in the udeb now?
[15:48] <hallyn> right
[15:48] <hallyn> whereas in the version we have in trusty, it was in syslinux
[15:49] <hallyn> or maybe syslinux-common
[15:52] <jamespage> hallyn, the isolinux package has that file
[15:52] <jamespage> albeit in ./usr/lib/ISOLINUX/isolinux.bin
[15:52] <jamespage> but I think you can add that via debian/rules to ipxe build
[15:53] <hallyn> jamespage: what the...  I looked there before and saw others, but not isolinux.bin
[15:54] <jamespage> hallyn, we just need to transition and update the ipxe build process - I expect that will be accepted into ipxe with a suitable patch for debian
[15:54] <hallyn> jamespage: I'm sorry, I swear I looked there.
[15:55] <hallyn> so yeah that should be trivial to fix
[15:55] <jamespage> hallyn, hey - np - that's why we ask each other for peer review after all!
[15:55] <jamespage> (so we don't go make ourselves look stooopid somewhere else :-))
[15:56] <hallyn> so meanwhile lemme go look at ipxe upstream
[16:20] <jamespage> rbasak, did we ever get to the bottom of why mongodb is stuck inproposed?
[16:20] <jamespage>     * i386: mongodb, python-loofah does not hint at anything to me
[16:24] <rbasak> jamespage: oh, I didn't realise it was. I'll take a look in a minute.
[16:28] <jamespage> rbasak, I think it might be because I dropped the mongodb-dev package
[16:29] <rbasak> Oh, I remember now.
[16:29]  * rbasak updates his utopic-proposed apt cache
[16:31] <rbasak> jamespage: mongodb depends on mongodb-dev, but src:mongodb no longer produces it.
[16:31] <jamespage> rbasak, doh!
[16:31] <rbasak> jamespage: so that makes mongodb itself uninstallable. That's the first part of * i386: mongodb
[16:32] <rbasak> jamespage: python-loofah depends on mongodb, so also becomes uninstallable. I expect that'll fix itself as soon as mongodb itself is fixed.
[16:32] <jamespage> rbasak, lemme sort out my mess
[17:21] <klep> when using ubuntu 12.04 with the latest MAAS and I'm having and issue with DHCP leases being handed out after the nodes are provisioned ... they never get DHCP once they are supposed to have started
[17:21] <klep> any ideas on this one?
[17:34] <rbasak> klep: maybe try #maas if you don't get an answer here?
[18:05] <hallyn> jamespage: corrected ipxe pushed to utopic;  will open a bug with jessie after I'm sure the utopic version is ok
[18:15] <jamespage> rbasak, oh great - https://launchpad.net/ubuntu/+source/mongodb/1:2.6.3-0ubuntu4
[19:23] <iclebyte2> is any one familar with dovecot? i'm having serious difficultly rebuilding a produciton server - i keep getting the error "SSL_read() failed: error:14094418:SSL routines:SSL3_READ_BYTES:tlsv1 alert unknown ca: SSL alert number 48,"
[19:23] <iclebyte2> and " Error: net_connect_unix(auth-worker) in directory /run/dovecot failed: Permission denied (euid=2000(vmail) egid=1002(vmail) missing +r perm: /run/dovecot/auth-worker, dir owned by 0:0 mode=0755)"
[19:24] <Sling> iclebyte2: what does nami -mo /run/dovecot/auth-worker show?
[19:24] <Sling> sorry, namei
[19:26] <iclebyte2> http://pastebin.com/HusqYszY
[19:28] <Sling> hm that socket should be nobody:nobody and 660 I think
[19:29] <Sling> try chown nobody:nobody /run/dovecot/auth-worker; chmod 660 /run/dovecot/auth-worker
[19:30] <Sling> or you seem to be running it as vmail, is this some shared hosting platform?
[19:31] <iclebyte2> yes it's hosting business. UK2.net messed up my server and i'm rebuilding it from scratch but ubuntu 14.04 is using dovecot 2.29 - my old box was 10.04 and i'm seriously struggling to get it going again
[19:31] <iclebyte2> lots of angry customers..
[20:47] <crashcourse> i'm getting 'mail transport unavailable' messages on my postfix server but with no other clues in the logs - postconf output is here: http://pastebin.com/P32BtEKj - any ideas where to start debugging this?
[20:55] <phunyguy> crashcourse: this channel is pretty inactive most of the time, especially with very specific issues like that.  Maybe #postfix would be able to help more...
[20:55] <crashcourse> cool, thanks
[20:56] <phunyguy> they are really good in that channel.
[20:59] <phunyguy> crashcourse: Also, that pastebin says you are on Debian... not Ubuntu
[21:05] <RoyK> phunyguy: not much difference...
[21:05] <phunyguy> RoyK: not the point.
[21:06] <phunyguy> RoyK: I wouldn't go into #debian asking about ubuntu questions.... just doesn't seem right the other direction either.
[21:36] <EL3PHANTEN> Hello Does anyone here know what I should do to log cookies. I have put this in my apache2.conf   ->  LogFormat "%{name}C" cookielog  and then  CustomLog /var/log/apache2/my-access.log cookielog... But All I get in the file is 127.0.0.1 - - [12/Aug/2014:23:13:50 +0200] "OPTIONS * HTTP/1.0" 200 - "-" "Apache/2.2.22 (Debian) (internal dummy connection)" "-"
[21:47] <justizin_> anyone know if there is an open bug for hash sum mismatch errors with 12.04.5 server iso? after fresh install apt-get update results in a slew of Hash Sum mismatch
[21:47] <sarnold> justizin_: which mirror?
[21:47] <sarnold> justizin_: are you using apt-cacher-ng by chance?
[22:21] <weaver> I'm tracking problems with the Ubuntu EC2 us-east-1 APT mirror, check sum mismatch, is anyone else seeing this?
[22:22] <weaver> *hash sum mismatch
[22:22] <sarnold> weaver: someone else reported some hash sum mismatches but didn't report the server :/
[22:23] <sarnold> weaver: are you using apt-cacher-ng by any chance?
[22:23] <weaver> no caching
[22:23] <weaver> W: Failed to fetch bzip2:/var/lib/apt/lists/partial/us-east-1.ec2.archive.ubuntu.com_ubuntu_dists_precise-updates_universe_source_Sources  Hash Sum mismatch
[22:25] <weaver> this error pops up regularly due to the S3 mirror update procedure not being transactional, but it typically resolves itself in minutes. We've been observing this for about an hour I think
[22:25] <sarnold> weaver: thanks, I've poked admins..
[22:27] <weaver> thanks sarnold
[23:41] <stevek> Hi all.  Ubuntu's PHP session cleanup script keeps trashing my servers despite having file-based sessions disabled
[23:41] <stevek> I've improved the session cleanup script to exit 0; if file-based sessions are not enabled
[23:41] <stevek> where can I submit this improvement?
[23:42] <sarnold> stevek: dpkg -S /path/to/cleanup/script  --- then ubuntu-bug <packagename> for whichever package owns the file
[23:42] <stevek> ah, cool, thanks