=== JanC_ is now known as JanC [01:26] I'm trying to pxeboot ubuntu 14.04. The installer starts, but then it seems to look in 'trusty-updates/restricted/debian-installer/binary-amd64/Packages and fails [01:27] It fails because that path does not lead to anything. There's just an ubuntu 14 iso dropped hosted on anther server and that file is not there. Why does it keep looking there? [01:29] Ok. all the other mirrors seem to have this except the one i set up locally. I think there's something wrong locally... === Milady is now known as Countess === Countess is now known as Milady === Milady is now known as Countess [06:19] jamespage: as of this night OVS master got the DPDK 16.07 bits [06:19] jamespage: 0a0f39df: netdev-dpdk: Add support for DPDK 16.07 [07:51] Good morning. [09:49] hi ... [09:49] im having error with all-in-one installer on 14.04.5 [09:49] fenris@bigsoftware:~$ sudo openstack-install [09:49] Traceback (most recent call last): [09:49] File "/usr/bin/openstack-install", line 282, in [09:49] ev = EventLoop(ui, cfg, logger) [09:49] File "/usr/share/openstack/cloudinstall/ev.py", line 43, in __init__ [09:49] self.loop = self._build_loop() [09:49] File "/usr/share/openstack/cloudinstall/ev.py", line 66, in _build_loop [09:49] event_loop=urwid.AsyncioEventLoop(loop=evl), **additional_opts) [09:49] AttributeError: 'module' object has no attribute 'AsyncioEventLoop' [10:30] Please use pastebin for console output. [10:32] ejat: http://askubuntu.com/a/722167 [12:34] Hello [12:34] I'm new to OpenStack and I would love to have it running in my own testlab at home [12:34] I've got 6 servers with 16-64GB Ram [12:35] Can I follow this guide to setup the master server: https://help.ubuntu.com/lts/clouddocs/installer/en/single-install.html and then add compute instances ( the other servers )? [13:23] rbasak: just saw that we got surpassed on pacemaker :-) [13:23] rbasak: anyway this was blocking an SRU which is ok now no matter which merge got in, could you accept the Xenial task in bug 1595901 for me? [13:24] bug 1595901 in pacemaker (Ubuntu) "Missing dependency on dbus" [High,Fix released] https://launchpad.net/bugs/1595901 === m1dnight1 is now known as m1dnight_ [13:44] rbasak: the bug now also got the relatively trivial SRU debdiff for Xenial [13:54] cpaelzer, re-snapshotting master and rebuilding against 16.07 dpdk [14:53] Hi can anyone help, for some reason my sites are giving an intermittent white screen of death in the browser?? [14:53] php error. [15:12] jamespage: thanks for the info [15:41] cpaelzer, amd64 is looking ok - doing some py3 enablement at the same time as this update [15:41] i386 - https://bugs.launchpad.net/ubuntu/+source/openvswitch/+bug/1609846 [15:41] Launchpad bug 1609846 in openvswitch (Ubuntu) "unit test failures with master branch on i386" [Undecided,New] [15:45] jamespage: did you upload that to the git-packaging repo of ovs already ? [15:45] just so I can try to find what is going on to help if I find some time? [15:45] cpaelzer, no [15:46] just pushed to git repo under ~ubuntu-server-dev [15:47] jamespage: thanks will pick up there then [16:04] I'm working on finishing the run once Default + crypted LVM test case for 14.04.5 right now [16:42] (for i386 fwiw) [17:57] Does anyone know if DDClient can be used for selective subdomains? e.g. wiki.domain.tld and blog.domain.tld on dynamic dns, and not mention forum.domain.tld because its A record points to a fixed IP. [18:11] beisner, python-keystoneauth 2.4.1-1ubuntu0.16.04.1~cloud0 is ready for promotion to mitaka-updates [18:31] Hello friends. I am trying to figure out how restore a TAR backup of an Ubuntu system to a freshly created VMware system (to test the integrity and process of restoring backups). I need some advice, all the way from partitioning, to reconfiguring GRUB. [18:39] coreycb, unable to do uca activities atm, querying in fn #launchpad as it looks like a redirect/rewrite change in behavior or something along those lines. [18:40] beisner, yikes, ok let's let that get sorted out. [18:41] coreycb, tldr; our tooling now fails with 404 on the .changes files [18:42] beisner, that smells like a launchpad outage [18:44] coreycb, well it's odd. check this: [18:44] our script reaches out to open: [18:44] https://launchpad.net/~ubuntu-cloud-archive-private/+archive/ubuntu/mitaka-proposed/+files/python-keystoneauth1_2.4.1-1ubuntu0.16.04.1~cloud0_source.changes [18:44] but gets 404 not found. i see the same with wget. however, if i hit a browser with that, i end up at: [18:44] https://launchpadlibrarian.net/276333938/python-keystoneauth1_2.4.1-1ubuntu0.16.04.1~cloud0_source.changes [18:44] which does load. it's as if there is a new redir/rewrite behavior that's less than polite. [18:46] beisner, that is odd. I get the same behavior fwiw [18:54] Hey guys, I am trying to restore a backed up Ubuntu system in TAR to a new server. Here's the details. The old system was set up with "Guided with LVM" and the target system has two partition, one at the beginning in ext4 and then an extended swap partition (No LVM). Once I untar the backup to the ext4 partition and grub-install then boot, it boots me into maintenance mode. What should I do to repair the restore partition in order to get it to boot pr [18:54] operly? [18:55] riz0n: not a realistic goal [18:56] technically possible but a lot of effort [18:56] riz0n: you should run update-grub at the least. and you you definitely need to update /etc/fstab with your new UUID and lack of lvm [18:56] riz0n: it might be easier to backup just the data you care about and then automate the deployment process instead [18:56] ikonia: What do you mean not realistic? How would you recommend that I backup an Ubuntu server then restore the backup to a separate machine? [18:57] riz0n: or if you want to continue with full-system backups, maybe investigate a full-system backup solution [18:57] using restore products [18:57] you've just untar'd old / not-working configs onto a new system [18:57] ok let me check out the /etc/fstab .... how would I go about finding out uuids? [18:57] the fstab the grub config [18:57] thats all going to be wrong [18:57] then there are more complex risks [18:57] riz0n: blkid [18:58] Thank you so much. I got the backup/restore process to work one time, but I used some utility called "boot-restore" or something like that. But I never could get it to work twice. [18:58] but there will be more as others are saying that I am not remembering at the moment [18:59] boot-restore is the least of your problbmes [19:00] oh and you will need to update-initramfs as that will be looking for lvm too [19:15] Figured I would give you guys an update. So editing /etc/fstab then update-grub fixed it! You guys are great, thank you so much for your help! [19:15] be sure to test all aspects of that system :) I'm still skeptical of a tar backup being Good Enuogh [19:17] I will, right now the backup consists of a fresh ubuntu install (yes I know, kind of crazy to "restore" a clean ubuntu install, but the goal is to hammer out the process necessary of backing up a system then restoring to a different in case of failure). === ashleyd is now known as ashd === petevg is now known as petevg_vacation === petevg_vacation is now known as petevg_afk === petevg_afk is now known as petevg_vacation [22:18] Heya folks, this is what I have been doing for the last few hours to get MaaS 2.0 up and running on 16.4.1? LTS, using four machines, all of which are the same apart from the server/controller/rack region controller having double the RAM of the nodes, this could do with some work, spellinga and grammar, but it's quick and dirty, and works.. https://ubuntuforums.org/showthread.php?t=2332842 [22:18] Now I need to find a way to install, Autopilot, OpenStack and Juju without any issues :D [22:18] I have also found that if sometimes, your node does not fully start, it does a partial "thing" where it can fetch the data required through a PXE boot and then like hang for ages..., reboot the node and switch it's UTP patch cable from one port to another empty one and it fires up proper first time... I think it's cheap switches, mines from TPLinnk and it still does it.. [22:19] Can anyone point me in the right direction of going about installing "autopilot" on Ubuntu 16.4.1LTS with Maas2.0 installed, please :D [22:21] ^^Get in, get out... Shake it baud about... [22:21] derwood2: seen this yet? http://www.ubuntu.com/download/cloud/install-openstack-with-autopilot [22:22] (I'm thinking mostly of step 6) [22:22] Let me just have a look, cheers for the link :D (I have seen a few links, and they more on about 14.4LTS, I've had some serious issues) brb :D [22:23] awesome :D brb "step 6" [22:23] yes, now that 16.04.1 is out, I hope more of the documentation can be updated for it.. [22:23] Me too, more documention is really required :D [22:25] Do you know what would be nice, if I could make a complete .ISO file from my current exact install of 16.4.1LTS with MaaS2.0 all sorted and configured... And then after I try this and it could break, I can just re-install using the nice one I've just spent two hours setting up.. [22:26] Okay. Noobie question... I need to start xserver, vnc server, autologin, and run a program we need to run... all with ubuntu server. brain hurting, anyone ever do this before? [22:26] I use blend [22:26] oops [22:27] let me start again, I use Blender, and I have cenotOS installed as a minimum, it will boot and login for me and start blender without even the normal desktop being seen, but the blender window.. [22:28] Does that sound like something you would want, but rather then blender, using VNC.. [22:28] yeah similar [22:28] funny [22:28] we are doing blender renders [22:28] ;) [22:29] does blender not provide a daemon for rendering? o_O [22:29] It has a very horrible one haha [22:30] Blender is awesome, if you search for a erm, help page on how to network render, a few of them tell you how to get the system to autologin, and start blender, ready for netrendering.. this is done to a node so it takes up the least amount of RAM and time to setup.. [22:30] We use loki [22:30] You could change the setting rather for blenders use, but for VNC and whatever else you need :D [22:30] Can you send us that ami? [22:30] it just seems .. silly and insane .. that something as routine as rendering would require a graphical environment [22:31] We are using many 3rd part apps to manage files and render managment [22:31] why on earth would they require such an annoyance? that would make it a thousand times harder to use on amazon or GCE or whatever [22:31] I'm actually right now, setting up a private cloud RACK cluster using MaaS2.0, Autopilot, OpenStack and Juju so that I can use the whole cluster to spawn a VM with blender in and use all avaliable RAM and CPU for rendering :D [22:31] GUI isnt required for blender we are doing some other crap with blender which needs a ui [22:32] I will use it also for computations and the like too :D [22:32] but thats the ease, or so I see it with MaaS, I can just config as I go, strip and redeploy depending on my needs. [22:32] Can you send me the ami? [22:33] I'm just trying to get it all to work, on this link, https://ubuntuforums.org/showthread.php?t=2327868 [22:33] derwood2: .. I'm afraid that you may have an incorrect idea about what is possible with openstack; do you think that you're going to get to launch a single VM with 64 gigs of RAM and eight sockets of CPUs? [22:33] Sure I'll find you the link for making blender autologin and start-up :D brb [22:34] in the beginning yes, you are correct I was thinking that would be the case, but as I went along I found I would have to create a set of VM's one for each bare metal chassis :( Shame eh.. [22:35] so confused hehe [22:35] derwood2: yeah; the old 'beowulf cluster' project existed to try to build Bigger Computers from little ones, but I don't know if anyone still uses it any more [22:36] I used to use Beowolf's way back in late 99/02 with MPI proper nice.. sadly I have been out the game for over 10 years.. [22:37] David... This might help you out I think : https://cgcookie.com/tutorial/setting-up-a-renderfarm/ [22:38] Oh, only issue right now is ubuntu server doing an autologin and starting a vnc server [22:39] david_theory: you could probably isntall ubuntu-desktop or unity, then use the control panel's "users" page to enable automatic login [22:39] it's a bit of a shame, that's a few hundred megabytes of RAM that'd be profitably used for rendering [23:09] I was hoping I could use MaaS2.0 to spawn a very small ubuntu install that would just contain the programs I require, even thought I could get them to spawn a VM with mpirun already contained rather then having to install all the nodes with a small copy. Like we used to.. This week of playing with MaaS is not looking to be a good use of time :( [23:10] I've just tried OpenStack again using section 6 and I've had to scrape it, I can get it to start to show with conjure-up? But some serious bootstrap errors using all the avaliable options like Nova and the other two. [23:11] I need to find out why this is happening, but I'm not sure how?