[00:03] anyone know how to enable the php hash plugin/functionality? [00:06] install uh [00:06] apt-get install php5-mcrypt [00:06] but I believe it already has some builtin to core [00:06] like md5, sha256 etc... === markthomas is now known as markthomas|away === peter is now known as Guest62005 [02:07] Hmm, where does do-release-upgrade (or is it just apt?) get its expected download speed from? I'm doing a 12.04 -> 14.04 upgrade right now, and it told me "You have to download a total of 992 M. This download will take about 1 minute with your connection." I mean, I WISH our connection was that fast ;) [02:10] keithzg: you are on 1Gbps ethernet, right ? :) [02:10] keithzg: erm 10Gbps [02:11] keithzg: alternatively, it might be going by how long it took to update the package indices, I'm not sure - have never bothered looking [02:12] has there been another shellshock patch? [02:12] google engineers said the things still super vulnerable to where it was before [02:12] in a different way [02:14] mdev: the hardening patch that we released on saturday should mitigate the scope of the remaining flaws down from "crisis" to "bug that needs to be fixed" [02:15] alrighty thanks [02:15] i'd say it'd be nice to avoid this sort of thing in the future but apparently bash is worked on literally 1 guy so... [02:15] yeah [02:16] and he's done heroic work this last week [02:16] thought an ubuntu dev fixed it for ubuntu? [02:19] yes, mdeslaur provided us with updated packages; he used Chet's patches :) [02:19] ahh nice [02:19] sarnold: for the record, I used redhat's patch, as did debian [02:20] Chet's patch came out a few hours later, and used a different suffix [02:21] mdeslaur: aha :) [02:21] and now apparently apple has used a different prefix/suffix combo also [02:21] well inital ubunut patch wasn't completely, was that from redhat? [02:21] >() or something.. go figure. [02:27] anyone know how I can lookup what version of ubuntu I'm running from initramfs prompt? [02:28] I know it is 12.04 but dont remember if there was anything beyond that [02:28] lsb_release -a [02:28] may have to apt-get install lsb_release [02:28] doesnt work from initramfs prompt [02:29] I have a broken array so loads into funky prompt [02:29] trying to get a boot usb going so I can boot into root [02:29] cat /etc/*-release [02:29] but want to make sure I get the right version [02:29] cat /proc/version [02:30] uname -a [02:31] hmm [02:31] cat /proc/version and uname -a both worked [02:31] but unexpected results [02:32] Linux (none) 3.2.0-43-generic #68-Ubuntu SMP [02:32] is what it returned [02:32] and a date [02:33] uname -r [02:34] same result less date [02:34] 3.2.0-43generic [02:34] must be because of the failed array and not being able to boot into root [02:34] lsb_release really the best way, when I first started porting code to nix I was rather annoyed it was so hard to cross distro version info reliabily like you can on windows, but apparently that's not really cared about in nix community [02:35] *get rather [02:35] you maybe able to get version info based off your repo but I don't know how === rcj is now known as Guest3347 === bilde2910|away is now known as bilde2910 [03:18] anyone have any experience with rebuilding broken arrays? [03:47] anyone have any experience with rebuilding broken arrays? === phunyguy is now known as phunyguy1 === phunyguy1 is now known as phunyguy [06:28] Good morning. [06:59] <|\n> hey guys, noob situation here... was trying to set it with a value much lower than 65535 and got "sysctl: setting key "net.core.somaxconn": Invalid argument" what should i read to get understanding on why i see this error at all? === ggherdov`___ is now known as ggherdov` [07:34] |\n: EINVAL for that sysctl should only be caused by the value being outside the valid range. [07:34] |\n: ..which is 0-65535. [07:34] <|\n> soren, i can not agree with that due to what i see, many thanks for pointing that out [07:35] |\n: How are you setting the value? [07:35] <|\n> soren, i've edited sysctl.conf directly and -p [07:36] <|\n> that is not all, i also got the same for net.ipv4.tcp_max_tw_buckets, net.ipv4.tcp_tw_reuse and most of things i tried to set out of experimental purposes [07:36] |\n: pastebin your sysctl.conf. [07:37] <|\n> k [07:38] <|\n> soren, http://pastebin.ovrnet.ru/paste/gH1BsuJ-#J2z2Ic2i [07:40] |\n: If you stop putting comments after the values, you'll be fine. [07:40] |\n: Those aren't comments. They'll get written to the sysctl, too, so you're not even passing in integeres. [07:40] integers. [07:40] <|\n> soren, oh, hah, good if so! many thanks! [07:40] np [07:41] <|\n> soren, i don't do such things often enough to stop myself from commenting things like that before i put em [07:41] <|\n> thank you, man [07:42] Sure. [07:54] rbasak, there is a bug for that writable /usr/lib thing in mysql already I think [07:54] let me check [07:54] morning soren [07:54] jamespage: 'morning! [09:27] jamespage, I've updated https://code.launchpad.net/~gnuoy/nova/bug1314677/+merge/236321 as per your comments however the patch has been -2 upstream due to stable/icehouse freeze for 2014.1.3 [09:27] gnuoy, bah - nm - we can carry it as a patch for .3 === markthomas|away is now known as markthomas === Lcawte|Away is now known as Lcawte [11:34] Hi. I'm running 14.04 server in a VMWare Datacenter (don't know exact which, but I'm using vSphere to access it). I have problems with this build server where it becomes mindbuggingly slow. It is doing dpkg-builds and may take ages to complete. We have a similar server running 12.04 on the same datacenter, with the same setup which runs fine. When the 12.04 takes 1.5 hrs to build all... [11:34] ...software, the 14.04 server may take 8-10 hours. [11:35] I can't figure out why this happens. From the data center side, the two servers are identical. Looking from the 14.04 guest side, the kernel log sais nothing, but you see it running at 100% on all cores and is generally very slow [11:36] I'm running the latest kernel 3.13.0-36-generic 64-bit. The core of the problem is: How can I figure out if this is a datacenter issue or a kernel/guest issue? [11:37] I'm lost and IT cant help me any more as I can't provide any description to what is wrong [11:43] zul, ok - what do we have rc's for - keystone? [11:43] jamespage: ? [11:44] jamespage: afaik nothing has changed packaging wise for keystone or glance [11:44] zul, yes - I mean't which upstream rc's [11:44] jamespage: glance and keystone so far [11:47] zul, http://pad.ubuntu.com/juno-rc1 [11:47] for tracking [11:48] jamespage: coolio...ill keep my eye on things and update it if i see any more [11:52] Hello, I was wondering why "w" was returning an ip address like 1.128.31.14. Ubuntu 12.04.05 kernel 3.8.0-44-generic. Is there a bugtrack I can check? [11:54] rtfmoz: Nothing you've said so far indicates a bug [11:54] ok what should I check... [11:55] Nothing, as nothing appears to be wrong? [11:55] zul, rage watch files [11:56] oh? my ip address is 101.168.42.151 [12:00] system seems to think I am coming from that address... 21:59:12.243307 IP 172.31.0.23.22 > 1.128.31.140.62793: Flags [P.], seq 1178096:1178304, but i dont think I am. will be back [12:00] to check something [12:02] 172.31.0.23 is a RFC1918 private address [12:02] Clearly NAT is involved [12:05] yes but thats happens inbound to that address. my source is internet address and should not be modified.. investigating. [12:11] zul, coreycb: hmm - either of you doing MIR for python-glance-store ? [12:11] jamespage, I can, the sync just went through yesterday [12:11] coreycb, please - its needed for glance rc1 [12:11] coreycb, I'll upload with it added to the deps [12:11] jamespage, ok [12:12] ok it seems 1.128.31.14 is a valid IP address after all. go figure. [12:13] inetnum: 1.128.0.0 - 1.159.255.255 [12:13] netname: TELSTRAINTERNET49-AU [12:13] coreycb, ok - just dealing with an issue in the unit test execution in glance [12:13] yep that the carrier I am using. just thought it was a bogus address. thanks === bilde2910 is now known as bilde2910|away [12:24] coreycb, zul: glance needs a minor bump on retrying as well - syncing that now [12:30] I am now testing server performance with sysbench, and I am seeing that my 14.04 server is on cpu performance is a factor 5(!) worse performance than on a twin 12.04 server installation. [12:31] I don't understand this. Kernel issue? Kernel on vmware host issue? Only on 14.04? [12:35] sveinse, are you benchmarking ontop of vmware? === bilde2910|away is now known as bilde2910 [12:37] jamespage, https://bugs.launchpad.net/ubuntu/+bug/1375770 [12:37] Launchpad bug 1375770 in ubuntu "[MIR] python-glance-store" [Undecided,New] [12:37] jamespage: Yes. Because this is the (corporate) environment we have to run our build servers. [12:38] jamespage: the 12.04 server performs adequate, but the 14.04 server performs very poorly for unknown reason. Takes factor 3-5 longer time to compile same software as on 12.04 server [12:39] sveinse, are they on the same underlying hypervisor? [12:39] jamespage: They are on the same host, with the same back-end storage solution [12:39] well, didn't the scheduler change? [12:39] for the disk? [12:40] did you tune any of them? or both just default installs? [12:40] both servers are running out-of-box solution for storage, that is with lvm2 default. No encryption [12:40] lvm2 default? that was never a default when I installed my servers [12:41] you need to check for differences [12:41] are both vm's built using the same scsi controller? same queue scheduler? same readahead? ..... [12:42] patdk-wk, yes, that is what IT is telling me. [12:42] heh? it? what is an it? [12:43] The biggest difference I can see is the kernel, but I guess you guys would have known if there were issues with the 3.13.0 kernel... [12:43] IT = IT department. The BOFH that controls the mandatory datacenter in our company [12:44] And they are giving up claiming that this is the 14.04 guest's fault, not the host system. So here I am [12:45] Well, you could install the trusty kernel on 12.04 and see if it causes the problem [12:46] qman__: What about the oposite? Taking the 12.04 kernel to the 14.04 server? [12:46] how exactly are you using sysbench? [12:46] sveinse, that would never work [12:47] patdk-wk: Rudimentary really. sysbench --test=cpu --cpu-max-prime=20000 run [12:47] The current kernel is backported to 12.04 and supported, the reverse is not [12:47] what? that is a cpu test, not disk [12:47] why would you think disk is the issue [12:47] That's the only reason I suggested it [12:47] Takes 30.2s on 12.04, while it uses 168.5s on 14.04 [12:48] but that doesn't even involve the kernel [12:48] I never said disk perf, did I? I said system performance. My apps compile awfully slow [12:49] ah [12:49] Disk performance is not that skewed, just a factor of 30-40%, instead of 500% percent difference on CPU [12:49] for me [12:50] 23.3525s on 14.04 [12:50] 23.7733s on 12.04 [12:50] so it is defently the 14.04 kernels issue [12:50] not that cpu tests even use the kernel [12:52] patdk-wk, on my VirtualBox 14.04 instance: 25.2s [12:53] sveinse, the 14.04 kernel does have more in-tree support for the esx hypervisor - its possible something there is causing issues; are you running open-vm-tools? [12:54] most likely, using a cpu's that lack ept support [12:55] jamespage: I was reccommended to install the commercial vmware tools from the hypervisor instead of using the open-vm-tools. Could this help perhaps? [12:56] sveinse, well its something to try [12:56] jamespage, it's still just cpu benchmark [12:56] tools aren't going make a difference [12:56] except to help network/disk/memory [12:56] hmm true [12:57] I would look for lack of ept support on the cpu, the host overloaded, or cpu overtemp [12:57] jamespage: I notice in vsphere that my 14.04 vm is setup with "hardware virtualization" disabled, and "performance counters" disabled. They are so for the 12.04, but could this be related as well? [12:57] what cpu is in this server? [12:57] patdk-wm: I would too, except 12.04 and 14.04 run on the same host server [12:58] does the server only have a single cpu? [12:58] sveinse: I am such a BOFH ;) [12:58] could be improperly installed heatsink on one [12:58] and intel is suprising resilent about mixmatched cpu's :) [12:59] sveinse: this is vmware, right? is vmware tools installed on the guest? I'd recommend using the vmware tools from vmware, not the open source variant [12:59] patsk-wk: I don't know the specs for it (as I don't have the rights to access this info in vsphere). I think its a 12-core AMD blade server (but I'm guessing) [13:00] RoyK: Yes, that was the recommendation from the datacenter admins as well [13:00] cat /proc/cpuinfo, should get you *close* [13:00] patdk-wk: on the guest? ;) [13:00] yes [13:00] it normally shows the *correct* cpu model name [13:00] it should show the cpu model, yes [13:00] but not the correct flags :) [13:00] but only the one the system was booted on, incase of vmotions it will be wrong [13:01] patdk-wk: AMD Opteron 6238 === bilde2910 is now known as bilde2910|away [13:01] coreycb, http://pad.ubuntu.com/juno-rc1 [13:01] for rc1 tracking - I'm dealing with keystone and glance - next is all yours ;-) [13:02] RoyK, you have any experience with running 14.04 on vmware data center? (unfortunately I don't which and which version) [13:03] sveinse, all of my tests where from ubuntu 12.04/14.04 running on esxi 5.5 [13:03] sveinse: I have a few VMs running 14.04, yes [13:03] patdk-wk: Thanks === Guest3347 is now known as rcj === rcj is now known as Guest90094 [13:05] jamespage, ah thanks! it looks like only Keystone and Glance have been released thus far? [13:06] What I'm seeing in vsphere management is that my 14.04 server is almost always maxing out the CPU figure (and gives CPU usage alarms). The other's are only a few 100's MHz, but 14.04 is consistently maxing out to 10320 Mhz (it is setup with 2 cores 2 sockets) === Guest90094 is now known as rcjenn [13:08] sveinse, "hardware virtualization" disabled sounds bad to me === rcjenn is now known as rcj` [13:10] jamespage, zul: could one of you merge this? https://code.launchpad.net/~cjohnston/horizon/icehouse-1308651/+merge/235978 [13:10] jamespage: I'll try to change that. I also see that CPU/MMU virtualization is set to automatic. What is it for you guys? [13:11] jamespage, zul: and this please, https://code.launchpad.net/~corey.bryant/horizon/b3-css-fix/+merge/236161 [13:17] sveinse, you want automatic [13:17] but where is hardware virtualization even an option? [13:19] In vsphere (the web page). My VMs are running VM version 10 (ESXi 5.5 and later) and as such only available on the web interface, not the dedicated client [13:19] Under CPU options [13:20] ah, I only use the web interface if required [13:20] vmware has changed the policy, so some new features are only available in the web interface [13:20] that has always been the policy [13:21] Setting HW virtualization enabled didn't help. 12.04: 30.3s, rebooted 14.04: 182.1s [13:23] This is just so weird [13:24] Hmm. Hang on. My 14.04 is set to 32-bit linux in the host, while running 64-bit linux. Could this be the cause? [13:25] it will cause issues, yes [13:25] cause using 32bit on a 64bit system is slower than crap [13:25] so it will use software virtualization for 32bit guests [13:25] instead of hardware [13:27] really? [13:27] * tafa2 did not know that [13:28] from #ubuntu [14:27:00] BluesKaj: kodiak1, kvm and qemu, virtualbox [13:28] patdk-wk: That's it! Just rebooted. And it took 10s to come online with ssh. Results: 28.8s :D [13:28] Hey folks, Redhat sysadmin here - wondering if Ubuntu has a supported oVirt-based virtualization product (like Redhat has with RHEV) [13:28] I am really happy guys! Let me test this for a day or two, but I am very optimistic [13:29] tafa2: Those aren't really a reasonable competitor to VMWare. If those are 'it', I'm guessing Canonical aren't really trying to compete with Redhat or VMWare in the virtualization market... [13:30] kodiak1 im just the messenger someone posted that after u left [13:30] tafa2, thanks haha [13:30] :) [13:30] what is ovirt? isn't that just the windows driver for kvm? [13:30] i thought that was virtio [13:31] ah, yes, virtio. I was confused. [13:31] https://en.wikipedia.org/wiki/Red_Hat_Enterprise_Virtualization is just KVM with extra stuff. [13:32] oh... this ovirt: https://en.wikipedia.org/wiki/OVirt huh, never heard of it. [13:33] kodiak1: I think you are correct in saying that Canonical has nothing which directly matches up to those products. [13:34] looks cool [13:34] "RHEV is just KVM with extra stuff" is kind of an understatement. [13:35] That's kind of like saying that Openstack is just KVM with extra stuff [13:35] kodiak1: i'm crazy enough, I probably would say that :) [13:35] I know that Canonical is heavily invested in Openstack but I'm still kind of surprised that they don't have the resources to roll a supported oVirt product. [13:35] jrwren: heh - kvm is just the hypervisor - there's a *lot* more [13:35] kodiak1: I know enough about openstack to know how foolish that would be. Is RHEV as much of a beast as OS? [13:36] If you've ever used vSphere & ESX in a clustered config you're seeing basically the gamut of what oVirt does [13:36] kodiak1: what would such a product do? Allow management of virtual machines? [13:36] kodiak1: transparent migration too? [13:36] All of it [13:36] kodiak1: that is impressive. good job redhat! [13:37] People like to say that Openstack is for managing herds and RHEV / oVirt is for managing pets [13:38] If you have specific long-running VMs that make up important long-term services they usually go on RHEV or VMWare, whereas appdev, testing, and more transient VMs go in Openstack [13:38] -generally- [13:38] That is one approach. Another approach is everything in openstack. [13:38] People are starting to use Openstack where they would traditionally use VMWare but that's kind of overkill [13:39] I think some folks have a goal of making openstack so easy that it is not overkill. [13:40] oVirt is much more mature as well, I think a lot of mid-size shops are waiting on dust to settle on openstack because as-is it takes a lot more care and maintenance [13:41] i don't know that it takes a lot more care and maint. In fact, that was not my experience. It required very little maint. It does havea very steep learning curve. [13:42] There's the learning curve [13:42] There's the upgrades [13:42] kodiak1: at this point we are really talking business and markets, right? RHT wants to sell in that market right now and apparently canonical does not. [13:43] Trust me as someone who's dealt with both recently oVirt is much simpler for people who don't have dedicated ostack people - and it'll run for way longer than havana, icehouse, etc are supported [13:43] kodiak1: is there a free open source ovirt thing that one can play with? Can I install fedora and try out the tools you are talking about? [13:44] Yep oVirt is 100% free upstream project of the RHEV product [13:44] I'm not too bummed that Canonical isn't in on oVirt in a paid-support kind of way, just a little surprised. I'm a Redhat guy coping with a buddy working on Canonical stuff so it's not a big deal to me [13:45] http://resources.ovirt.org/pub/ovirt-3.4/ [13:45] kodiak1: :) Why be "a redhat guy" why not be a linux guy ? [13:46] When there are a million other things to stay current on it's easier to focus on what the employer is interested in using. [13:47] I do check in on Canonical every few years. I was a Debian user from 2002 on and used Ubuntu from 4.10-9.10 [13:47] kodiak1: always good to stay aware of what is out there. [13:47] Allegedly oVirt is supported on Debian [13:48] kodiak1: i saw that on ovirt.org. Probably not much to make it work on ubuntu. [13:48] coreycb, zul: keystone done [13:48] jamespage: sweet [13:55] coreycb, do you have anytime for https://code.launchpad.net/~gnuoy/nova/bug1314677/+merge/236321 ? [13:55] gnuoy, sure I'll take a look [13:55] thanks [14:04] If I have a UnixListenSocket at /var/chroot/dev/log, how do I redirect the input to it to /var/log/blah.log? [14:04] Within rsyslog, the only possibility is to create a socket, not where the input is directed [14:12] bug #1322568 also affects cloud-archive:icehouse, should this be added to the bug report and if yes, how? or will the updated libvirt packages be backported anyway? [14:12] Launchpad bug 1322568 in libvirt "nova interface-attach fails" [High,Fix released] https://launchpad.net/bugs/1322568 [14:19] coreycb: ceilometer is out [14:20] zul, thanks === mattgriffin is now known as mattgriffin-afk [14:38] I personally dislike vmware - a lot === peetaur2_ is now known as peetaur2 [14:44] tafa2: why? === mattgriffin-afk is now known as mattgriffin [14:44] RoyK so clunky [14:45] coreycb, both of those horizon merges merged; I switch the changelog entries to UNRELEASED - please do that for any changes that won't be uploaded immediately [14:45] much prefer for example onapp [14:45] tafa2: clunky? [14:45] jamespage, ok, and thanks! [14:45] jamespage, can we merge liam's update to nova? https://code.launchpad.net/~gnuoy/nova/bug1314677/+merge/236321 [14:45] coreycb, already done! [14:46] just call me Mr Merge [14:46] tafa2: rather different things, though [14:46] jamespage, whoa, Mr Lighting Fast Merge! [14:46] RoyK yes - and I meant it from a hosting perspective [14:47] :) [14:47] for that, I'd probably used openstack [14:49] jamespage, coreycb, thanks [14:49] gnuoy, yw - nice job [14:49] gnuoy, the UNRELEASED comment above is a good one for you as well btw [14:50] coreycb, gnuoy: using UNRELEASED means that additions to the changelog get appended to the UNRELEASED entry [14:50] jamespage, yep, I did update the changelog after your comment [14:50] coreycb, gnuoy: leave releasing to the uploader... normally zul or I but might be others [14:51] jamespage, k thanks for the tip [15:55] rbasak, or jamespage or anyone.. [15:55] o/ [15:55] owonder if we have a sane source of information that shows package upgrades between T->U [15:55] in a consolidated view (ie, not "go crawl utopic-changes") [15:55] What do you mean? [15:55] A list of packages that are newer? [15:56] yeah. and i guess new packages too. [15:56] For the entire archive? [15:56] and dropped. [15:56] well, i care about main more than universe [15:56] and universe more than multiverse [15:56] I'm not aware of one, but some shell-fu with grep-dctrl and comm or meld might be enough [15:56] but i assume if such a thing exists it probably has data for all. [15:56] Maybe distrowatch? [16:01] zul, coreycb: oslo.vmware 0.6.0 also in the release team review queue for glance rc1 [16:01] I'll get there even if it kills me [16:01] jamespage: okies [16:02] jamespage, thanks [16:03] oslo.vmware? why 'oslo'? it's where I live :P [16:03] open stack library something or other === kickinz1|afk is now known as kickinz1 [16:05] smoser: how about something like http://paste.ubuntu.com/8466864/ ? [16:05] (could do with some tuning though) [16:07] yeah. not unreasonable as a start. [16:07] this was for someone asking "what else is new in ubuntu server" [16:07] and me wanting a way to somewhat easily answer that. [16:08] I guess maybe we want to filter that to packages subscribe to in LP by ~ubuntu-server or something. [16:08] There might be stuff in universe that'd slip through that net though === Ursinha is now known as Ursinha-afk [16:16] zul, are you intending doing anything with ironic? its currently bit-rotted all release stuck in proposed [16:17] jamespage: not really, i thought someone else was taking care of that [16:17] zul, who? [16:17] jamespage: adam_g [16:18] oh === kickinz1 is now known as kickinz1|afk [16:53] <[1]Az> I want to ignore the error I get when I try and run an uninstalled program, but its not on STDERR [16:53] <[1]Az> is there a way around this? [16:54] [1]Az: 2>/dev/null [16:55] <[1]Az> ah [16:55] <[1]Az> i had a space in there like an idiot [16:55] <[1]Az> too pro [16:55] [1]Az: > somewhere or 1> somewhere sends stdout to somewhere, 2> somewhere sends stderr to somewhere [16:55] <[1]Az> yeah i know :) [16:55] 2>&1 sends stderr to stdout [16:55] <[1]Az> but I'd done 2 > /dev/null [16:55] ah [16:55] :) [16:56] <[1]Az> and assumed it was a kernel error [16:56] <[1]Az> which you cant send to /dev/null? [16:56] it very rarely is ;) [16:56] <[1]Az> dd does it i think [16:56] <[1]Az> oh no it wasnt dd [16:56] <[1]Az> it was something weird [16:56] <[1]Az> cant remember now [16:57] <[1]Az> one of time or sync has unpipable messages [17:26] you are reading more than me about it [17:26] lol [17:27] hallyn, So if you not beating me to it, I would do that forward of the libvirt/apparmor discussion tomorrow [17:29] smb: sure. reply go guido now though to tell him you'll do so [17:30] ack [17:36] smb: zul: you'r enot working on any utopic libvirt push right now are you? [17:36] hallyn: hell no [17:36] nope [17:37] ok, pushing a quick fix for bug 1375910 [17:37] Launchpad bug 1375910 in libvirt "package libvirt-bin 1.2.8-0ubuntu6 failed to install/upgrade: subprocess installed post-installation script returned error exit status 1" [High,In progress] https://launchpad.net/bugs/1375910 [17:50] I want logrotate to store the rotated files in a subdirectory [17:55] Hello. running 12.04 and am noticing this HWE business. Just trying to make sure i'm understanding the ramifications of updating to 12.04.5 HWE: Does this mean that i will be moving from a 3.5 kernel to a 3.13 kernel? [18:05] if you are on anything other than 3.2, yes, you will get 3.13 [18:05] wedgie: correct, this is the package you'd be using https://launchpad.net/ubuntu/+source/linux-lts-trusty [18:05] anything inbetween is *unsupported* [18:07] ok, thanks === Ursinha-afk is now known as Ursinha === bilde2910|away is now known as bilde2910 [19:19] anyone have any experience with rebuilding broken RAID arrays? I had a drive fail completely in an mdadm array 5 and now when I boot my server it boots into initramfs prompt. [19:20] when I choose the option to load a degraded array it seems to ignore my response and proceed to initramfs prompt anyway [19:28] running ubuntu server 12.04. looking for someone to help walk me thru adding a swap file on a system [19:28] swapon -s [19:28] Filename Type Size Used Priority [19:29] Device Boot Start End Blocks Id System [19:29] /dev/sda1 * 63 36812799 18406368+ 83 Linux [19:29] /dev/sda2 41082880 80021503 19469312 5 Extended [19:29] /dev/sda3 36812800 41082879 2135040 82 Linux swap / Solaris [19:29] /dev/sda5 41084928 80019455 19467264 83 Linux [19:29] so theres no swap on currently.. yet i have a partition available for it. [19:30] all the walkthroughs i've seen dont seem to cover this scenario [19:30] nyktovus: Could you pastebin your /etc/fstab? [19:30] !paste| nyktovus [19:30] nyktovus: 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. [19:33] http://paste.ubuntu.com/8468072/ [19:33] thanx for the link [19:38] nyktovus: Your swap partition is not defined in fstab that is why it is not used. [19:38] agreed. [19:39] so how should i best define this? [19:40] utlemming: heads-up, https://bugs.launchpad.net/ubuntu/+source/lxc/+bug/1373981 sounds like a bug in cloud images (awaiting more info) [19:40] Launchpad bug 1373981 in lxc "ubuntu 14.10 server, amd64 - lxc-start fails to start container" [Undecided,Incomplete] [19:40] nyktovus: Add an entry like "/dev/mapper/padme-swap_1 none swap sw 0 0" [19:40] nyktovus: Of course edit to your configuration. [19:41] how bout add " [19:41] /dev/sda3 swap swap defaults 0 0 [19:41] " [19:41] sorry.. there shouldnt have been line breaks in that [19:41] nyktovus: /dev/sda3 swap sw defaults 0 0 [19:42] nyktovus: /dev/sda3 none swap sw defaults 0 0 [19:42] Disregard the previous one ;) none needs to be added. [19:42] which one? [19:44] ok.. so i added your line.. reboot? or do i need to tell it to swap on or something? [19:46] nyktovus: Try 'mount -a' or 'swapon', I forgot... [19:47] swapon -s is now showing my listing [19:48] i dont need the UUID of the partition in fstab? [19:48] It's preferable but not mandatory. [19:48] why preferred? [19:48] hwat is preferable? the UUID of the partition or the UUID of the filesystem? [19:48] nyktovus: sdXY namings can change, UUID's should not. [19:49] gotcha. [19:50] apparantly i've been running this server for years with no swap... been flawless tho [19:52] nyktovus: should be like that [19:52] nyktovus: if not get more ram [19:54] anyone have any experience with rebuilding broken RAID arrays? I had a drive fail completely in an mdadm array 5 and now when I boot my server it boots into initramfs prompt. [19:54] anyone have any experience with rebuilding broken RAID arrays? I had a drive fail completely in an mdadm array 5 and now when I boot my server it boots into initramfs prompt. [19:54] oops, sorry for the double post [19:55] hey all i have a wierd question. I am running ubuntu server 14.04 and during boot up the resolution is fine once it gets to the login prompt the resolution is just a small box on the screen. any one ever experienced this before [19:55] so i shouldnt use a swap? [19:55] nyktovus: swap is nice but if you have too much swap traffic too often it's a sign you need more memory. :) [19:56] nyktovus: swap can be useful sometimes, to keep stuff in memory when it's _not_ used [19:56] oh i dont have that issue at all.. i'm at 20% usage as we speak.. [19:56] its on a dell 1907 monitor [19:58] slops17: this might be of some help http://askubuntu.com/questions/299975/proper-way-to-change-terminal-resolution-in-ubuntu-server-13-04 [19:58] iDealz, i tried that already and didnt work [20:02] anyone have any experience with rebuilding broken RAID arrays? I had a drive fail completely in an mdadm array 5 and now when I boot my server it boots into initramfs prompt. I believe I need to boot from a CD to get into root, but dont know where to go from there [20:15] iDealz, i found this and seams to have worked [20:16] http://www.thomas-krenn.com/en/wiki/Fixing_Graphics_Output_Driver_Problems_of_LES_with_newer_Linux_Kernels [20:20] slops17, good thing we don't do graphics in this channel :) [20:21] patdk-wk, i had problem number 3 on that link only the upper left portion of the screen was in use and it was a paint to do anything on the box but it is now fixed === markthomas is now known as markthomas|away === bilde2910 is now known as bilde2910|away === Ursinha is now known as Ursinha-afk [20:54] Howdy folks. I just updated my LXC binary, and now my apparmor service is dying on me with the below error. Can someone help me understand what "TOK_MODE" means so I can maybe take a stab at remediating it? [20:54] "AppArmor parser error for /etc/apparmor.d/lxc-containers in /etc/apparmor.d/abstractions/lxc/container-base at line 42: syntax error, unexpected TOK_OPENPAREN, expecting TOK_MODE" [20:58] Logos01: Hi - what Ubuntu release are you running? [20:59] Logos01: also, can you copy and paste line 42 of /etc/apparmor.d/abstractions/lxc/container-base? [20:59] 14.04 [21:00] err, 14.04.1 (just updated) -- Whole line is " unix (receive)," [21:00] Logos01: I have run into issues with apparmor scriptsa few months ago, don't know if they have been fixed, have stopped trying [21:00]