[00:00] skorv: connect to the CUPS admin webpage [00:01] Ah. No I am not able to run the examples. I get the error that the function does not exist, as before [00:08] Interesting. You have the same GD version as I do. When I run the example on the PHP.net website (http://www.php.net/imageantialias) I get the same error. === JamesHarrison` is now known as JamesHarrison [00:45] What software do the Ubuntu repos use for management? apt-mirror? [00:48] i'm having some trouble using ssh-agent. when i pidof ssh-agent, it seems to be running, but ssh-add -l claims not to be able to connect to the agent === squishy is now known as SquishyNotHere === SquishyNotHere is now known as squishy [01:08] StrangeCharm: that's because you're using screen [01:14] twb, interesting. can you tell me more? [01:18] Sorry, I'm busy just now [01:18] http://code.haskell.org/~twb/Preferences/.bin/twb-agents [01:18] Hmm, not much documentation there [01:19] http://code.haskell.org/~twb/Preferences/.bin/.bash_profile shows how I create it [01:19] twb, bottom line: i shouldn't have trouble if i just start a non-screen session? [01:20] also, your bash profile is a 4040 [01:21] Sorry, [01:21] http://code.haskell.org/~twb/Preferences/..bash_profile shows how I create it [01:21] ARGH [01:21] http://code.haskell.org/~twb/Preferences/.bash_profile [01:21] re short answer -- yes [01:21] The problem is that your screen session outlives your agent session, so screen still points to the one that existed LAST time you logged in [01:22] You need to either update environment variables (nontrivial), or start ssh-agent by hand [01:22] The gentoo keychain docs might explain more [01:26] so your bash-profile has a trick in it for keeping screen pointed at the right agent? [01:32] http://superuser.com/questions/141044/sharing-the-same-ssh-agent-among-multiple-login-sessions [03:09] hi guys. I'm trying to help someone with a windows network, who's experiencing periodic lock-up. they might be working, or simply open large files over the local LAN & their system would simply lock up (they describe it as "going off to la-la-land"). I've rebuild their Ubuntu server, running *only* SSH, Samba & Webmin, and *nothing* else. load it pretty good, so now I'm trying to find other possible causes or remedies (I've set === squishy is now known as SquishyNotHere === SquishyNotHere is now known as squishy [03:57] hi guys [03:57] anybody here ? [04:09] I'm here...you have a question? [04:10] yes hi [04:10] i am having issues installing ubuntu server 10.10 [04:10] am I in the right channel ? [04:11] yes [04:11] ok [04:11] i have downloaded the latest iso file from the official site itself [04:11] made a bootable usb key out of it [04:11] booted comp on it [04:11] followed steps normally [04:11] and then at some phase, the "Select and isntall software" phase [04:11] instlaler just hangs at 5% [04:12] for no apparent reason [04:12] tried to reboot and instlal in expert mode changing a couple of settings but nothing will do [04:13] Chrystopher: There is a log file during the install that may tell you more about what is going on. [04:13] Chrystopher: I would also try ubuntu 10.04 since it is a long term support unless you need something specific on 10.10 [04:14] i actually don't really [04:14] I'll try with the older version [04:14] it's much more stable or ? [04:14] Chrystopher: 10.10 is only support for 6 months...yes 10.04 server is supported for 5 years. [04:15] ok will try and let you know [04:15] 10.10 just has newer software [04:16] but 10.04 still has the same basic packages ? [04:16] yea for sure [04:17] ok I guess I could be good with that [04:17] I actually still run a ton of 8.04 and it works fine for me. [04:17] oh one thing though [04:17] Which software would you recommend to create bootable device ? [04:18] i've been using the one proposed by ubuntu [04:18] but i've heard unet bootin is good as well [04:18] usb-creator-gtk ? [04:18] I have used that a bunch. [04:18] oh ok [04:18] problem is i must do it on a windows machine though [04:21] ah I admit I haven't done it much on windows but if you are getting to the installer you are doing it correctly. [04:26] hi guys. I have a few disks I'm about to scratch. should I continue using ext4, or should I risk btrfs? [04:31] Chrystopher: sorry I was having tons of problems with my wireless...the last thing I typed was usb-creator-gtk [04:32] hm nope [04:32] the last thing you typed [04:32] concerned your lack of experience with windows [04:32] lack is a bit strong but.. anyway you get me [04:33] still donwloading [04:36] yea if you are getting to the installer it sounds like the install is going fine....try the 10.04 version and see what happens...if it hangs again I would start checking logs. [05:31] New bug: #704766 in nagios3 (main) "unable to install nagios3" [Undecided,New] https://launchpad.net/bugs/704766 [06:03] formatting live [06:07] theshedd17: are you still here ? [06:07] ff* [06:08] yea [06:09] if you type the username correctly I will be notified :) [06:09] because I'm usually have a bunch of stuff going on. [06:10] np === someone is now known as Guest13211 [06:10] Did 10.04 work? [06:10] it's weird there's either a problem with the software creating the usb key or the distro itself [06:10] I've loaded the key [06:11] now the only thing displaying is: "SYSLINUX 3.06 2010-04-01 EBIOS Copyright [...] " [06:11] and command prompt [06:11] but I can't type crap [06:11] and nothing's going on [06:11] lol [06:12] hmm..when it is booting? [06:13] yeah [06:13] well i manually select usb key to boot it from [06:13] and there it is.. [06:13] i will try with an alternative software [06:14] Chrystopher: how did you do the stick? [06:14] on windows with universal-usb-installer-1.8.2.5 [06:14] Chrystopher: I load up ubuntu desktop version X and use usb-creator-gtk :-/ [06:15] should i try to format it to make sure ? [06:15] Chrystopher: did you select ubuntu from list or other linux? [06:15] Chrystopher: can you link me to the web site you are using for windows I will take a look.. [06:15] Chrystopher: yea the usb-creator-gtk you select the iso and the usb drive and click create. :-/ [06:16] i followed the link to download the software form this page : http://www.ubuntu.com/server/get-ubuntu/download [06:16] it leads to this page: http://www.pendrivelinux.com/universal-usb-installer-easy-as-1-2-3/ [06:17] Chrystopher: if you didnt select ubuntu from dropdown but other linux it wont work [06:18] i will format and start again just to make sure [06:18] distro i choose is [06:18] Ubuntu Server 10.04.1 32bit installer [06:20] i mean from dropdown of installerä [06:20] i have noticed that if you dont select it from dropdown but select other linux it wont boot, just syslinux appears === Guest55686 is now known as lau === Kiall|AFK is now known as kiall === TeTeT_ is now known as TeTeT [11:37] anybody here uses libvirt with kvm? I got problems with virsh restore - 50% of change it will succed, 50% it will hang (it is logged as causing INFO: task kvm:6431 blocked for more than 120 seconds) [11:37] is this problem normal that there is no safe way to restore saved vm? === Cyber_Rock is now known as who === who is now known as croooo === croooo is now known as Cyber_Rock === oubiwann_ is now known as oubiwann === oubiwann is now known as oubiwann_ [13:45] <[diablo]> afternoon all... anyone running Ubuntu Server with rsyslog being used for centralized logging please? [13:45] me [13:46] <[diablo]> hi gobbe [13:46] <[diablo]> gobbe, did you go with RELP, TCP or UDP, etc please? [13:46] with syslog? [13:47] <[diablo]> yep [13:47] <[diablo]> well, rsyslog [13:48] http://www.google.fi/url?sa=t&source=web&cd=1&ved=0CBkQFjAA&url=http%3A%2F%2Fwww.ubuntu.com%2Fsystem%2Ffiles%2FCentralLogging-v4-20090901-03.pdf&rct=j&q=ubuntu%20rsyslog&ei=gOs2TbyeGcqfOpvI4JgE&usg=AFQjCNEL3sc8W3FDTyyFUyG4alXFQP8o2w [13:48] uuh [13:48] sorry [13:48] but however, that's pdf for rsyslog, quite good howto [13:48] i'm doing with [13:48] TCP [13:49] <[diablo]> yeah I am looking at that atm [13:49] RELP would be good also [13:49] <[diablo]> yep [13:49] <[diablo]> thats what interests me atm [13:49] <[diablo]> RELP [13:58] SpamapS: did you test the eglibc/upstart fix for maverick and lucid too? [14:01] Hello guys. For a big files download server, what's the most important part for being able to handle big traffic? CPU? RAM? [14:02] Fidelix: extremely high bandwidth limits, CPU, RAM, hard disk space, etc. [14:03] EvilPhoenix, but what does the webserver (nginx) use the most? CPU or RAM? [14:03] I've got space and I've got bandwidth. I'm concerned about the speed though [14:05] And being able to support all requests [14:06] fast disks is also quite important [14:06] so if you have lots of downloads disk-io is critical also [14:06] gobbe, good point... [14:07] But what I'm really concerned is if CPU is critical for this. I've got a relatively cheap server, and all its specs are good, except for the CPU. [14:09] i would say that cpu isn't so critical [14:14] Fidelix: I have a file server that has delivered about 5.5 million images today. CPU is not an issue on that box. [14:15] * zul grumbles [14:15] This is golden information. Thank you very much gobbe, andreasf [14:15] Fidelix: No problem :-) [14:15] andreasf, what web server do you use to serve these images? [14:15] * andreserl Morning!! [14:16] Fidelix: Apache, actually. [14:16] andreasf, well, that's surprising. Are you using a reverse proxy? Serving files through x-filesend ? [14:16] Fidelix: You might consider putting av caching server in front of your web server. Varnish, for instance. [14:17] andreasf, is that what you do? And does varnish help for big files? (by big, i mean, 60MB + [14:18] Fidelix: This particular server does not use a reverse proxy / cache, just Apache. If you use a Varnish server, images will be delivered from RAM, not disk. [14:18] Fidelix: Well, after they have been fetched from the backend. [14:19] heh, I totally don't get that [14:19] andreasf, that would be a good idea for small files i believe. But not for big files [14:19] wouldn't just putting more ram into the apache server do the same, havine more cache ram [14:19] Fidelix: Depends on how much ram you have :-) [14:19] andreasf, 32GB [14:20] Fidelix: You can configure Apache or Varnish to keep objects (images) in the cache for a specified amount of time, for instance 10 seconds or 10 hours. [14:20] But there are a lot of big files. They would fill 32GB in a second [14:21] <[diablo]> anyone know how to rsyslog to RELP? TCP and UDP are @ and @@ ... find no info on what clients use for RELP [14:23] Fidelix: Have a look here: http://www.varnish-cache.org/docs/2.1/ [14:27] is anyone here running a maverick desktop with virt-manager? [14:28] hallyn: i am [14:29] hallyn: you mean virt-manager in maverick or a maverick dekstop VM with virt-manager? [14:29] hallyn: this is server-channel, so desktop discussion should be on #ubuntu [14:30] lol [14:33] hallyn: i am, too [14:33] When I receive messages in pidgin, the message pops up in the top right corner unless the chat is active. How can I disable that ? [14:34] milligan: Have you asked in #ubuntu ? #ubuntu-server is really for Server questions (cli). [14:34] alright - my bad :) [14:35] :) [14:55] doko_: no I have not tested it on lucid or maverick yet. [14:56] SpamapS: could you do this? I prepared uploads. Maybe I should upload these to a PPA first === oubiwann_ is now known as oubiwann [15:05] gr, started a vm-clone before remembering the nbd hang in current kernel. now i have to reboot. [15:10] doko_: Yes I'll test in the next 2 hours in VMs. [15:10] doko_: and thanks! [15:11] doko_: if you were to upload yours to a ppa, that would certainly make it easier for me to do that testing. [15:11] SpamapS: will take more than 2h to build [15:11] doko_: do you have a branch I can just branch/bzr bd ? [15:11] SpamapS: so if you have a faster machine, I can point you to the sources instead [15:11] SpamapS: no [15:13] doko_: right.. eglibc... since we're just changing the maintainer scripts.. does it allow building with nocheck? [15:13] I think it takes my machine about an hour to build eglibc [15:14] SpamapS: yes [15:15] doko_: ok well if you can point me to the eglibc and upstart sources I'll start the build before I have to run an errand that will take about an hour. :) [15:17] SpamapS: see chinstrap:~doko [15:18] kirkland: nice interview....and nice props to the team ;) [15:18] http://www.talkincloud.com/ubuntu-server-and-the-cloud-notes-from-a-developer/ [15:19] robbiew: thanks, this author did a good job recapping the interview [15:19] (seems like too often an article like this gets it all wrong -- not the case here!) [15:21] yeah...and sometimes they spin it completely wrong...glad that didn't happen [15:21] whew! [15:24] doko_: sorry I don't know how chinstrap works.. help? [15:25] SpamapS: copied to http://people.canonical.com/~doko/tmp/ [15:26] doko_: ack, building [15:27] SpamapS: not the ~ppa version [15:28] Daviey SpamapS: so the cobbler "pristine" source doesnt use the Makefile so i been scrambling a bit more [15:32] doko_: right, 2.11.1-0ubuntu7.8 is building now in a lucid chroot [15:32] kirkland: howdy!! Should I create a new powernap-common package that installs the class that manages the config, plus the Monitors? (not that powernapd daemon instantiates this class) [15:33] Are there cloud AMI's for 10.10 [15:34] elb0w: http://uec-images.ubuntu.com/maverick/current/ [15:34] ty [15:38] andreserl: yes, definitely [15:38] andreserl: also, let's bump the version to 2.x [15:39] andreserl: these are significant new features [15:39] kirkland: ok cool :) [15:40] kirkland: will finish updating the manpages then. But if you wanna test before I finish, go ahead since the packaging has already been updated to install the new stuff [15:43] andreserl: cool [15:46] SpamapS Daviey: ok uploaded again [16:21] hello [16:22] I am trying to install avamar agento on my ubuntu server. It is asking for libcap1 which I cannot find in apt-get. Any thoughts? [16:26] says libcap1 is not available. Package libcap1 has no installation candidate? [16:27] hi all. I have an ubuntu server that seems to be in trouble - it restarted twice today without anything useful in the logs except two lines that I don't understand http://pastebin.com/yALsFcYE [16:28] knixy: libcap2 - support for getting/setting POSIX.1e capabilities [16:28] knixy: iirc libcap2 is the one used on ubuntu [16:28] knixy: libcap1 is quite old [16:28] ok, let me try again [16:29] knixy: if avamar needs libcap1, I guess finding a pre-built lib somewhere is the way to go [16:29] * SpamapS returns [16:29] hmm.. eglibc still building [16:29] * RoyK flees [16:30] * SpamapS pursues [16:30] RoyK: are you running proprietary display drivers?-) [16:30] JamesPage: it seems d-i changed again [16:30] this is what I get when I try to install the avamar deb [16:30] gobbe: just 10.04 server - no fancy stuff except for those two 8-core CPUs and 64gigs of memory [16:30] hggdh: whats the impact? [16:30] http://pastebin.com/wFtuRGw6 [16:31] JamesPage: all tests failed; running one local I see d-i stopping and asking about what type of partition I want (Sun, msdos, AIX, etc) [16:31] RoyK: ok, kernel taint happens usually with weird drivers :) [16:31] RoyK: have you run memory checks? [16:32] knixy: unless there's a lot of data or specific services, I would recommend using nfs for this one [16:32] JamesPage: it seems my saving of d-i log is now functional (see http://204.236.234.12/job/AKT-DBench/25/artifact/25/test-results/) [16:32] gobbe: not yet - I'm not at the office - but I just can't beleive I get those taint messages now - nothing's changed [16:32] RoyK, ok. Debian is supported for avamar, but ubuntu is not, which makes me sad :p [16:32] gobbe: it's basically a compute node... [16:33] knixy: it's about time EMC wakes up :) [16:33] damn right [16:33] * RoyK uses bacula [16:33] RoyK: yeah, that's just something, but i have seen several cases where kernel taint's are coming with display or wireless drivers [16:33] hggdh: excellent; looking at the daily tests I think xango3 is suffering from its file permissions issue on the iso images [16:33] we use avamar on all our servers here, 70+, if I don't get this working, then I have to switch the server to redhat., sigh [16:34] gobbe: it should at least have told me which driver was loaded [16:34] JamesPage: the change (where the preseed stops) can be seen here: http://pastebin.ubuntu.com/555844/ [16:34] or _what_ tainted the kernel [16:34] RoyK: that's true [16:34] JamesPage: yes, I need to grab the current libvirt and rebuild it with the permission patch [16:34] * hggdh goes to get it done [16:35] hggdh: I had some local success on maverick using the /etc/libvirt/qemu.conf settings [16:36] JamesPage: setting 'dynamic_ownership=0' ? [16:36] mdeslaur: could you please take a look at bug #525287 [16:36] Launchpad bug 525287 in lvm2 "Add support for corosync based clusters in clvm" [High,Confirmed] https://launchpad.net/bugs/525287 [16:36] hggdh: yes - I also configured it to run as root [16:38] JamesPage: hum. Will try it again -- I remember setting dyn_own to zero, but I do not remember is onwer was root, and it did not work [16:38] andreserl: please ask someone from the server team...zul? [16:38] nope...foundations i dont want to touch lvm2 [16:38] mdeslaur: ^^ [16:38] zul: who do you think I should contact from foundations? [16:39] hggdh: yes - I think that you need the change to root otherwise the normal libvirt account can't access the files r/w [16:39] hggdh: root just trounces everything :-) [16:39] andreserl: colin maybe [16:39] JamesPage: All hail root :-) [16:39] what is the cmd to find out how to remove avamar? [16:39] trying now [16:39] JamesPage: I will try to zero in this debian-installer change [16:40] knixy: apt-get remove? [16:40] zul: this would be clustering...who on the server team does clustering? [16:40] hggdh: of course I would never recommend doing this on a piece of production infrastructure but hey - its my laptop :-) [16:40] mdeslaur: me!! [16:40] zul: ok thanks [16:40] gobbe, im not sure the whole name. isnt there a grep cmd to find out the exact name of it? [16:40] JamesPage: exactly my view ;-) [16:40] andreserl: hehe, sorry about that :) [16:40] mdeslaur: no probs :) just changed nicks btw, I'm RoAkSoAx :) [16:41] hggdh: not sure about the d-i issue; the one test that ran on mercury AM today did pass.... [16:41] andreserl: ah! that would explain my confusion :) [16:41] oooohhh andreserl is RoAkSoAx... [16:41] * hggdh jots it down [16:41] mdeslaur: yeah... it indeed is confusing and hggdh just proved lol! [16:41] hggdh: xD === RoAkSoAx is now known as andreserls === andreserl is now known as RoAkSoAx === andreserls is now known as andreserl [16:44] andreserl: yeah, we should select nicks that anyone can pronounce [16:45] * hggdh disregards the implicit irony of previous statement [16:45] buona sera [16:45] !list [16:45] This is not a file sharing channel (or network); be sure to read the channel topic. If you're looking for information about me, type « /msg ubottu !bot » [16:45] hggdh, lol!! that's why I indeed changed to this nick... though noone ever recognizes me with it lol [16:45] heh [16:45] knixy: dpkg-query [16:45] well i fixed it [16:45] lol, i was alien a redhat rpm to deb [16:46] so I got the real debian deb, and booom. installed [16:46] thanks guys! [16:50] zul, hmm.. looking at the get-orig-source stanza, how come you switched back? [16:51] Daviey: because the new get-orig-source generates the pristine tarball [16:51] any linux wiz around that might help me understand these? http://pastebin.com/yALsFcYE - it's a server system, no graphics card or anything needing fancy drivers, and it happened during run time, server rebooted half an hour later for reasons unknown..... [16:55] zul, Aye, i just wondered if you worked out what was wrong with the other way? [16:55] Daviey: dont know really [16:55] Daviey: im happy with it the way it is now [16:57] zul: do I see that we're actually adding cobbler4j back in? [16:57] SpamapS: er yep [16:57] zul: maybe the cobbler project publishes it with different artifacts? [16:57] SpamapS; nope i checked [16:58] zul: so it just sits in git and rots? :( [16:58] SpamapS: pretty much [16:58] zul: well I'm glad we're including it. :) [16:58] will suggest then that they 1) remove the jar from git, and 2) publish the rest in the sdist. [17:00] royk, http://fonality.com/trixbox/wiki/kernel-tainted [17:00] patdk-wk: I know, but nothing happened on the box except the normal runs during that time - no su/sudo - nothing [17:01] that page looks to be saying, probably hardware issue [17:01] hello people [17:01] but need more logs to know what [17:01] i just recently installed webmin , and tried to login with root [17:01] but now i got "Error - Access denied for 127.0.1.1" [17:02] 127.0.1.1? [17:02] thats an odd address [17:02] why? it's perfectly valid [17:02] well thats error what i got [17:02] lo by default has a /8 netmask [17:03] dont you mean 127.0.0.1 ? [17:03] so 127.x will work :) [17:03] man i copied it [17:04] its 127.0.1.1 [17:04] patdk-wk: there are no more logs - nothing more than those two lines - but I guess it may be a hardware issue [17:06] royk, this seems to be the only thing that can cause a hardware issue, that sets taint [17:06] http://en.wikipedia.org/wiki/Machine_Check_Exception [17:07] andreserl: you are the cluster expert right? [17:07] dunno :( [17:07] cole, I'd not consider myself an expert, but highly knowledgeable, yes :) [17:08] Build needed 01:34:19, 1775336k disc space [17:08] * SpamapS fires up a fresh test VM [17:08] btw, does anybody know how to remove vms from virt-manager? [17:09] kind of annoying that you can create, but not destroy, from it [17:09] virsh undefine... [17:09] cole, what can I help you with? [17:09] SpamapS: I use virsh [17:09] they're not listed in virsh [17:09] list --all [17:10] SpamapS: right click, delete is an option [17:10] oh thats annoying [17:10] why would they not show up? [17:10] destroy server [17:10] hallyn: its not on mine [17:10] undefine server [17:10] SpamapS, is the VM stopped? [17:10] SpamapS: then connect to the one where they are (virsh -c qemu:///etc) [17:10] wait [17:10] now it is [17:10] or maybe it was.. [17:10] and my hatred of hiding everything in context menus prevented me from finding it [17:11] SpamapS, lol you have to shutdown the VM before being able to delete it :) [17:11] you were doing virsh list --all to show the ones that are down? [17:11] I think I was just being blind [17:11] k [17:11] forget everything I've said since "btw, does anybody know how" [17:11] :) [17:12] i'm good at amnesia [17:12] SpamapS, Dallas left you blind man? Too much golden hair xD [17:12] ! [17:12] * SpamapS still seeing Stars after two nights near the Lemon Bar ;) [17:12] SpamapS, +1 [17:13] andreserl: would there be any value in adding crm to keepalive pkg or does it only talk to services provided by pacemaker? [17:13] andreserl: poorly stated, obviously meant adding crm deb as a dep for keepalive === cmagina is now known as cmagina-lunch [17:14] jdstrand: i'm trynig to reproduce your kvm instability on ecryptfs, fwiw. (no luck yet, but trying) [17:14] cole, pacemaker and keepalived are two totally different things. In fact, keepalived can be seen as some kind of Resource Manager/Messaging layer, such as would pacemaker/corosync or pacemaker/heartbeat are [17:15] smoser: this is quite awesome: http://ubuntu-smoser.blogspot.com/2011/01/failsafe-and-manual-management-of.html [17:15] also, hi folks [17:15] jiboumans: o/ [17:16] its less valuable than it used to be. if you're using ebs root, catastrophe can be fixed. [17:16] smoser: the awesomeness is in the upgrades + failsafe together [17:16] jiboumans: o/ [17:16] smoser: [17:16] its there, almost. [17:16] did you see that ? [17:16] RoAkSoAx: hey, and belated congrats [17:17] smoser: not in details [17:17] the latest images can be launched with grub-pv kernel, they just don't do it by default. [17:17] nice [17:17] where can i RTFM? [17:17] and the latest dailies use it by default. [17:17] really only i my release announcement [17:17] smoser: the awesomeness is in the [17:17] oops [17:17] https://lists.ubuntu.com/archives/ubuntu-cloud/2010-December/000466.html [17:17] paste fail [17:19] smoser: awesome. safe enough to launch a production service with you reckon? [17:20] jiboumans: thank you!! [17:21] what are you guys using for web interface? [17:21] like webmin [17:21] is there anything else? [17:21] maudI use that [17:21] jiboumans, yeah, its solid. i would recommend using the pv-grub kernels. [17:21] sorry, i use webmin [17:21] just curious [17:22] smoser: nice. are you switching default aki's at some point you think? [17:22] the idea is to do that in next round.. the dailies do it now. [17:23] jdstrand: oh, but i'm NOT using encrypted filenames. you're using ext4 right? [17:24] hallyn: I am using ext3 for /home, with encrypted home [17:25] smoser: looking forward to it.. it's an awesome feature for us and i'm sure for many more folks on aws [17:25] smoser++ === Steve[cug]_ is now known as Steve[cug] [17:28] jdstrand: d'oh, i'm testing ext4, but that likely won't matter. but are you using encrypted filenames? [17:32] hallyn: I am [17:33] jdstrand: hm, well all 3 vm's appear to be hung. i didn't witness any segfaults, but they won't budge [17:33] hallyn: try logging into a console [17:34] hallyn: I saw something similar, and logging into a console and trying to run some commands showed things were really wrong [17:34] jdstrand: i had previously installed ubuntu-desktop, which is also hung [17:35] hallyn: it could just be networking-- I've seen it as somewhat flaky occassionally [17:35] * jdstrand wonders if it is virtio net... [17:35] but gdm is hung in all 3, so i can't log in [17:35] hallyn: I meant on tty1 [17:36] right but i can't get to tty1 [17:36] hallyn: but gdm hung is definitely interesting [17:36] well no, not gdm per se, [17:36] the whoel vnc session [17:36] so having gvncviewer send ctrl-alt-f1 does nothing either [17:36] hallyn: I see. can you disconnect the vnc and reconnect? what about sshing in? [17:37] no route to host - but i was ssh'd in when they hung [17:37] i had been disconnected from vnc, so each time i'm reconnecting [17:38] hm, [17:38] i'm out of space on the ecryptfs partition (8K to spare) [17:38] drat [17:38] ah [17:38] that could do it [17:39] I was always able to vnc, fwiw [17:39]