[04:16] <prgCoder> hi all, I am trying to compile some Embedded SQL/C programs on ubuntu server 12.04 and I keep getting the following error:
[04:16]  * genii twitches on !pastebin
[04:17] <prgCoder> libq.1.so: undefined reference to 'IIGchkobj'
[04:17] <prgCoder> ans some others - any ideas ?
[05:12] <lifeless> hallyn: hey - https://bugs.launchpad.net/nova/+bug/1228977
[05:12] <lifeless> hallyn: its targeted to saucy, but doesn't appear to have had the changeset pulled in
[05:13] <lifeless> hallyn: there seems to be some procedural issue blocking it
[05:35] <hallyn> lifeless: yeah it needed the sru justification.  i was waiting on a test case.  thanks for providing it :)
[05:35] <lifeless> we'll see about running a test with it for you
[05:35] <lifeless> hallyn: do you need to update teh build in proposed to include the security updates?
[05:35] <hallyn> great, thanks.
[05:36] <hallyn> hm, maybe.  i didn't see the security updates go by
[05:37] <hallyn> i only see an 	1.1.1-0ubuntu8.5, so should be ok
[05:38] <hallyn> gotta run - ttyl
[06:08] <lifeless> hallyn: ciao
[06:21] <mastershake> hey guys im getting a "The system network service is not compatible with this version" error, can anyone lend a hand for a moment?
[06:32] <mastershake> anyone?
[06:33] <sarnold> mastershake: is that an exact quote or a guess?
[06:34] <mastershake> sarnold: just a question... i keep getting a "The system network service is not compatible with this version" error
[06:34] <mastershake> and i dont know how to correct it
[06:37] <sarnold> hrm, I can't find that string in the debian code search
[06:43] <mastershake> you dont have to be rude about it
[06:43] <sarnold> mastershake: sorry, it's not rude, just stating that my primary tool, source code, isn't much help here :/
[07:38] <hxm> hi, why is this? http://apaste.info/6Fb5
[07:40] <hxm> changed resolv
[08:23] <zetheroo> is there a way to stop jbd2 from running on certain disks?
[08:35] <Bitwise> Hello. I've set up postfix, dovecot, and roundcube, seemingly unsuccessfully. When I send an email using roundcube, the recipient never receives it. Same result with result with telnet. I'm thinking there is a problem with the postfix configuration but I really haven't changed much.
[08:37] <mardraum> Bitwise: read mail.log
[08:38] <Bitwise> mardraum, Touche. I tried to send an email to my gmail account: (Host or domain name not found. Name service error for name=gmail.com type=MX: Host not found, try again)
[08:44] <mardraum> sounds like dns is broken
[08:52] <Bitwise> I did postqueue -f and now it says network is unreachable and connection timed out.
[08:54] <mardraum> is your dns working
[08:55] <Bitwise> Well I'm connected and browsing the net fine right now.
[08:55] <mardraum> from the mail server?
[08:56] <Bitwise> I'm running the mail server on my primary machine.
[08:56] <mardraum> I don't know what that means
[08:58] <Bitwise> Yes, I'm browsing the web from my mail server.
[08:59] <Bitwise> Are the first two lines correct? http://pastebin.com/S15yrCGn
[09:50] <zetheroo> is there a way to stop jbd2 from running on certain disks?
[11:25] <love12>  guys i have squid3 installled in debian  but there is aproblem when i try to test from the serever"The system returned: (111) Connection refused"
[11:27] <ikonia> so #debian would be the right channel to talk about that in
[11:27] <ikonia> you can join that channel with "/join #debian"
[12:02] <zetheroo> is it possible to stop jbd2 from running on certain disks?
[12:05] <ikonia> zetheroo: it's a kenrel process, as I understand it, it needs to access each disk
[12:05] <ikonia> zetheroo: you can tune it for each disk/file system though
[12:05] <zetheroo> ikonia: it's very unwanted for disks on which we have live VM images ;)
[12:06] <ikonia> it may well be possible, but I think you'll have to "tune it" not to run, as opposed to disable it
[12:06] <ikonia> if you get what I mean
[12:06] <zetheroo> ok ... kinda ...
[12:07] <zetheroo> how does one go about "tuning" it?
[12:07] <ikonia> hdparm
[12:08] <zetheroo> hmmm ... I have only ever used hdparm for modifying the spin-down settings
[12:09] <ikonia> to be honest, the same here
[12:09] <ikonia> well, and a few other very minor settings
[12:13] <zetheroo> isn't jbd2 the journaling process?
[12:15] <zetheroo> I am trying to find how hdparm can be used to "tune" jbd2 ... all I see is how to use it to change the spin-down settings ...
[13:36] <wiherek> hi
[13:36] <wiherek> I am setting up a vpn, on Ubuntu
[13:37] <wiherek> when I set up the localip, should I keep my current ip?
[13:37] <wiherek> or can it be some other value?
[14:29] <roadmr> Hello ubuntu-server people. We're looking at removing checkbox (checkbox-cli package) from the ubuntu server image for several reasons. I'd like to ask if anybody has objections about this (or really any other comment about it)
[14:35] <jrwren> its already not in cloudimg, so I'm all for it :)
[14:35] <roadmr> jrwren: hehe :)
[14:42] <zul> roadmr:  why?
[14:53] <roadmr> zul: the candidate replacement has some packaging issues so we're looking at options to avoid delivering a bad experience
[14:53] <zul> roadmr:  ok cool
[14:54] <roadmr> zul: removing it is the "easy way out" but we may have a solution to keep it if people would still prefer to have it
[14:54] <zul> roadmr: im ambivalent to it :)
[14:55] <roadmr> zul: cool, feedback appreciated :)
[14:55] <rbasak> roadmr: maybe email the ubuntu-server mailing list? I don't think anybody will object, but we should make some reasonable effort to check with any stakeholders. Are there any other suitable lists?
[14:56] <roadmr> rbasak: maybe maas would be interested. The mailing list idea is good, but if you don't see an e-mail from me, it means we found a way to keep it alive :)
[14:57] <rbasak> OK, sounds good :)
[16:56] <hallyn> rbasak: what would be the best way to snapshot a uvt-kvm vm?
[16:58] <RoyK> hallyn: what's uvt?
[16:58] <rbasak> hallyn: however libvirt suggests doing it, so with virsh directly. I tried the other day though, and couldn't get it working. Pretty sure it's not libvirt-specific.
[16:59] <rbasak> hallyn: if it's awkward we can have uvtool wrap it, but I need to know how to do it with libvirt first :)
[16:59] <hallyn> RoyK: uvtool (see package in trusty)  it lets you quickly create/use ubuntu-cloud-iamge-based libvirt vms
[16:59]  * rbasak has been writing documentation this week
[16:59] <RoyK> k
[16:59] <hallyn> RoyK: http://s3hh.wordpress.com/2013/12/12/quickly-run-ubuntu-cloud-images-locally-using-uvtool/
[17:00] <RoyK> hi all. I have eth0 on int 19, and all interrupts are processed by core0, which is getting rather hot with >10k interrupts/s. I've set smp_affinity to 0f, having 4 cores, but it's still stuck at core 0. any idea why?
[17:00] <rbasak> Manpages specifically. Server guide later.
[17:00] <hallyn> rbaska: sounds good, thanks
[17:00] <hallyn> gah.  rbasak: ^
[17:00] <jrwren> that is awesome, thanks for that.
[17:00] <jrwren> i'd been doing the virsh create manually
[17:01] <hallyn> RoyK: might ask on #ubuntu-kernel...  i would expect smp_affinity to DTRT...  might be a bug
[17:01] <RoyK> DTRT?
[17:02] <jrwren> would be sweet if that uvt-kvm had a way to pass user-data to cloud-init
[17:02] <rbasak> jrwren: --user-data :)
[17:02] <rbasak> (sorry the manpage isn't written yet!)
[17:02] <jrwren> rbasak: a file?
[17:02] <rbasak> jrwren: yes, or there are options to add common things for you automatically
[17:02] <jrwren> excellent, thanks.
[17:03] <rbasak> eg. --ssh-public-key-file, --packages, --password, --run-script-once, etc.
[17:04] <jrwren> this is great. how easy is it to bzr branch lp:uvtool and run out of trunk?
[17:04] <rbasak> jrwren: ppa:uvtool-dev/trunk :)
[17:04] <jrwren> i mean I might want to patch.
[17:04] <rbasak> I run PYTHONPATH=. python bin/uvt-kvm. That works from the source tree.
[18:36] <foolhardy> after clean install, ubuntu server x32 hangs on boot at "fsck from util-linux"
[18:37] <foolhardy> strangely, going into recovery mode and selecting "resume normal boot" bypasses the issue
[18:37] <foolhardy> any idea how to fix it permanently
[18:37] <foolhardy> ?
[18:38] <foolhardy> http://i.imgur.com/EGN3uiu.png
[19:20] <foolhardy> fresh install of ubuntu server 32bit. first boot this happens: http://i.imgur.com/nJTv3ru.png and then http://i.imgur.com/EGN3uiu.png
[19:20] <foolhardy> any idea why?
[19:21] <RoyK> nothing else?
[19:21] <sarnold> first boot looks normal
[19:21] <sarnold> or, rather, nothin in the first boot screenshot looks surprising to me
[19:25] <jrwren> http://www.ubuntu.com/getubuntu/releasenotes no 12.04.3 or 12.04.4 notes
[19:25] <foolhardy> I was worried that the ureadahead main process (246) terminated ... was a problem
[19:25] <sarnold> foolhardy: naah, ureadahead is always complaining about something. I just uninstall it.
[19:26] <foolhardy> it hangs here: http://i.imgur.com/EGN3uiu.png
[19:26] <foolhardy> never moves on
[19:26] <sarnold> now that -is- an issue :/
[19:27] <RoyK> foolhardy: can you boot it in single?
[19:27] <foolhardy> single?
[19:27] <foolhardy> going into recover mode and then selecting "resume normal startup" bypasses the issue
[19:27] <RoyK> recovery mode == single user mode
[19:28] <foolhardy> gotcha
[19:28] <foolhardy> dpkg and fsck do not fix
[19:28] <RoyK> fsck -f?
[19:28] <foolhardy> haven't tried that, doing now
[19:28] <RoyK> dpkg has nothing to do with this
[19:28] <RoyK> don't fsck a mounted filesystem
[19:29] <RoyK> unless it's mounted read only
[19:29] <foolhardy> http://i.imgur.com/Tbzlcjt.png
[19:29] <RoyK> looks good
[19:29] <RoyK> foolhardy: any reason for using 32bit?
[19:30] <foolhardy> it is virtualized so in trying to use as little resources as possible this vm has 1gb ram
[19:30] <foolhardy> I read that 64bit with less than 4gb ram runs very slowly
[19:31] <bekks> foolhardy: Thats a lie.
[19:31] <jpds> 'very slowly' ? I've never had that problem.
[19:31] <foolhardy> no?
[19:31] <bekks> foolhardy: No.
[19:31] <foolhardy> I have another vm with 2gb running 64bit but it is live now so I cannot change/replace it. I've never noticed an issue with it but I'm not guru.
[19:31] <foolhardy> ok, then. I'll just use 64bit
[19:32] <foolhardy> thanks a lot for the info
[19:32] <sarnold> foolhardy: 64 bit in 'small' machines does waste some memory for the double-sized pointers, but the extra registers on the CPU and indirect addressing modes of the CPU are fantastic for processing speed
[19:32] <RoyK> foolhardy: you can use 512MB for a 64bit VM without issues
[19:32] <RoyK> just depends what you're running on it
[19:32] <foolhardy> tomcat server is it
[19:32] <RoyK> I've had VMs with as low as 128MB on 64bit
[19:32] <foolhardy>  /java
[19:32] <foolhardy> well, great. THanks a lot for the great info.
[19:33] <foolhardy> I love you.
[19:33] <sarnold> (though in all honesty, 32bit _should_ work better than this.)
[19:33] <foolhardy> this is 12.04 lts. I had the same issue with 12.04.3 about six months ago and thats why I went 64bit on the other "live" machine. Only later to read that this was bad with less than 4gb.
[19:34] <RoyK> sarnold: obviously, yes
[19:34] <foolhardy> But now I know better.
[19:34] <foolhardy> 64bit 4eva
[19:35] <RoyK> foolhardy: saving resources by running in 32bit might not be a good idea. the 32bit instruction set has fewer registers, so things can't be optimized that well
[19:38] <sarnold> at some point in the future ubuntu will have support for a mixed mode of 32bit addressing but with the full 64bit instruction set, which will provide -some- memory savings for small vm instances, but I can't imagine that it'd free up more than 3-5% of memory
[19:42] <bekks> Well, basically there is no need to ride the dead horse named "32bit" when having 64bit hardware.
[19:44] <RoyK> bekks: indeed
[19:45] <RoyK> standardising on a common platform just makes things easier
[19:45] <RoyK> not that everything should run AMD64, but still
[19:47] <bekks> 32bit is dead for more than a decade in almost all home computers - and it's artifically revived by 32bit netbooks.
[19:50] <RoyK> bekks: I don't think there are much 32bit netbooks anymore
[19:50] <RoyK> even atom has been 64bit for years
[19:51] <bekks> RoyK: Yeah do have 64bit atoms nowadays, but there are still zillions of 32bit netbooks in use out there.
[19:52] <RoyK> sure - I have one ;)
[19:52] <bekks> Me too :P
[19:52] <RoyK> but the netbooks more or less died out when the pads came around
[20:41] <paco11> hello masters!
[20:42]  * RoyK waves
[20:43] <paco11> i have 12.04 with kernel 3.8.0-35-generic x64. It's fully supported to install the "saucy" kernel? (linux-image-generic-lts-saucy|linux-image-3.11.0-17-generic). thanks very much!
[20:44] <RoyK> paco11: why do you need 3.11?
[20:45] <paco11> to avoid the performance problem with kernels 3.0 to 3.9
[20:52] <paco11> RoyK: what do you think?
[20:57] <RoyK> paco11: what are those performance issues?
[20:59] <paco11> write performance
[21:07] <rabbel> clear
[21:07] <bekks> paco11: Can you be more precise please? I havent noticed "the performance issue" yet, using Linux on a variety of boxes.
[21:10] <paco11> i have good write performance in my ldap servers with 2.6.32 than 3.8.0
[21:13] <bekks> I dont have any issue with 3.2.0, 3.5, 3.8 as well.
[21:22] <paco11> http://lwn.net/Articles/486311/ | http://lkml.indiana.edu/hypermail/linux/kernel/1309.1/01585.html | http://www.openldap.org/lists/openldap-technical/201309/msg00187.html
[21:33] <foolhardy> anyone here run ubuntu server in proxmox?
[21:38] <sarnold> heh, I saw the question before the nick, I was just about to say "hey foolhardy does..." :)
[21:38] <foolhardy> My issue from earlier apparently wasn't connected directly with the 32bit edition, nor a single proxmox node.    :/
[21:38] <sarnold> :/
[21:39] <foolhardy> I can duplicate the issue on two different machines
[21:41] <sponzor> hi. my server is having 100% rame usage 24/7 http://pastebin.com/gNktmvzN
[21:41] <sponzor> host is on esxi server
[21:42] <sponzor> the rest of the server have normal usage.. but this one having 100 all the time
[21:42] <sarnold> sponzor: what do the swap in and swap out columns of 'vmstat 1' look like?
[21:43] <hitsujiTMO> sponzor: its only using 50% ram there
[21:44] <hitsujiTMO> sponzor: read the -/+ buffers/cache line. thats how much ram is actually being used. the rest is just filesystem buffering which gets dumped the moment an app actually needs the ram
[21:46] <sarnold> sponzor: in general "unused ram is wasted ram" -- but if you're seeing swap traffic, that could be indicative of something worse
[21:46] <sarnold> sponzor: using two megabytes of cache seems fine on a first guess, though.
[21:48] <hitsujiTMO> also swap only starts getting used when you hit 60% of ram usage(thats the default vm.swappiness) ... sinces there's only 2mb of swap used, that does suggest you're only barely going over that 60% at most
[21:49] <sponzor> yeah but 8gb memory usage in esxi doesnt loke nice. can i release un used ram in ubuntu?
[21:49] <RoyK> sponzor:  it's 2MB out of 8GB
[21:50] <RoyK> sponzor: linux swaps out things not in use
[21:50] <RoyK> sponzor: it doesn't matter!
[21:50] <RoyK> sponzor: you can turn off swapping if you like, sysctl vm.swappiness = 1, but I wouldn't recommend it
[21:51] <sarnold> sponzor: then I suggest your esxi monitoring tools are busted
[21:51] <hitsujiTMO> sponzor: "doesn't look nice" is a rather naive excuse for trying to free up the filesystem buffers. You're wanting to break optimisations just to skew usage figures
[21:51] <RoyK> sarnold: esxi monitoring doesn't take into account what linux is using its memory for
[21:51] <sarnold> RoyK: figures
[21:52] <RoyK> sponzor: what sort of server is this?
[21:52] <sponzor> mail server zimbra
[21:52] <RoyK> hungry beast, zimbra
[21:52] <RoyK> but 8 gigs should do for most
[21:52] <RoyK> how many users?
[21:52] <hitsujiTMO> sponzor: all modern operating systems do this, windows, osx, unix, linux ... they all take advbantage of unused ram
[21:52] <sponzor> 40 users
[21:55] <RoyK> sponzor: check out the munin plugins for monitoring zimbra
[21:55] <RoyK> they're quite good
[21:58] <sponzor> will try that.. i see it can be integrated with nagios witch we alrady run
[21:59] <RoyK> munin plugins with nagios?
[22:00] <RoyK> I've ditched nagios - using icinga - but the plugin format is the same
[22:02] <sarnold> RoyK: any thoughts on shinken?
[22:03] <RoyK> sarnold: haven't heard of it before now
[22:03] <sarnold> ah :)
[22:07] <sponzor> did tryed icinga... almsot the same as nagios. but we stay with nagios for now.. have 9 servers in distributed environment.. so its a pain to switch.. if we had money for xi work would be so easyer :D
[22:08] <RoyK> 9 nagios boxes?
[22:08] <RoyK> how many units are you montoring?
[22:10] <sponzor> 50-200 per box
[22:11] <RoyK> ok
[22:11] <RoyK> may I ask where you work? around 1k boxes seems rather a lot
[22:11] <sponzor> this could be done with only one nagios server.. but locked networks. for security reasons we had to deploy nagiso servers to remote sites
[22:12] <RoyK> ic
[23:12] <jrwren> why would you not want to use every byte of ram you have ?
[23:24] <rostam> Hi I am using ubuntu 12.04, I have installed dhcp server on my system.  By default when system reboots, the dhcp server is enabled, I like to change the default behavior for boot time. How could i do that? thx
[23:27] <sarnold> rostam: echo manual > /etc/init/nameofservice.override  -- see the cookbook for details http://upstart.ubuntu.com/cookbook/#override-files
[23:29] <rostam> sarnold, thanks