[00:00] that is shell for libvirt [00:00] you can remove the default network [00:01] PryMar56, brctl delbr virbr0? [00:01] sure, till reboot [00:02] PryMar56, can I permanently do that (since I don't think I use it), or would I need to put something in /etc/init.d? [00:03] PryMar56, I do have VMs in kvm/libvirt, but they're using other bridges, none of them are using virbr0 [00:05] virsh net-undefine default [00:06] * PryMar56 apologizes for spoon feeding [00:08] PryMar56, thanks [00:09] when masquerading, what if do I specify, the one I'm expecting a connection on, or the one I'm connecting to? === MeltedLux is now known as MeltedDed [00:19] acovrig_, looking at the iptables trace, the bridges are MASQ'd [00:19] you can dump those rules as `iptables-save` === Lcawte is now known as Lcawte|Away [00:28] PryMar56, I'm currently double-NAT'd (for the next ~week) 192.168.5.0/24 is an ubuntu machine running pfsense as a router VM (libvirt); it connects (WAN) to a 192.168.2.0/24 LAN; if I have 'iptables -t nat -A POSTROUTING -o br0 -j MASQUERADE' in 5.4 (the VM host), other clients can't connect to anything on the 192.168.2.0/24 network. [00:29] PryMar56, 5.4 is a client-to-client-LAN gateway to the 5.0/24 network from an openvpn overlay network (to a VPS) [00:30] PryMar56, when setting this system as a gateway, do I MASQUERADE br0 (LAN with 5.0/24) or tun0 (OpenVPN client IP to overlay network)? [00:34] don't try that at home, setup is for professionals only... pretty complicated to say for me [00:50] Having trouble with postfix not recieving or sending outside domain [00:50] any takers [00:52] i can login to the pop server remotely and send and recoeve to only local accounts === markthomas is now known as markthomas|away === zz_DenBeiren is now known as DenBeiren [04:21] kirkland: are you there? === l2ksolkov___ is now known as l2ksolkov [04:32] hi [04:32] i got gpg error : link this [04:32] http://paste.ubuntu.com/9566355/ [04:33] how to delete gpg key or disable that === bilde2910|away is now known as bilde2910 === Lcawte|Away is now known as Lcawte === EuaD is now known as ubuntuaddicted === Lcawte is now known as Lcawte|Away [09:40] Good morning. === Lcawte|Away is now known as Lcawte [11:11] coreycb, zul: I've bumped new versions of alembic and oslo.utils into vivid; oslo.context is in the NEW queue [11:42] jamespage: whats using oslo.context? [11:43] zul, neutron [11:45] zul, does the py 2.7.9 incompatibility matertilize via unverified certs? [11:45] jamespage: it does === MeltedDed is now known as MeltedLux [11:45] zul, where is the bug again? keystone is impacted [11:45] jamespage: https://github.com/python/peps/blob/master/pep-0476.txt [11:46] jamespage: https://bugs.launchpad.net/cinder/+bug/1403068 [11:46] Launchpad bug 1403068 in cinder "Tests fail with python 2.7.9" [Undecided,New] [11:50] jamespage: i have a patch that ignores some test failures for cinder if you want (rtslib) [11:52] zul: I think glanceclient is impacted as well [11:52] fuuuu [11:58] zul: hmmm [11:58] yes [11:58] oslo.vmware 0.8.0 uploaded [12:02] zul, ok trying a patch to glance tests [12:02] glancelient that is [12:03] coreycb, btw I'm applygin a new uversionmangle - "s/\.0b/~b/" [12:04] jamespage: http://paste.ubuntu.com/9569176/ [12:09] zul, I see the issue with glanceleitn [12:59] zul, glanceclient fixed and proposed upstream [13:00] jamespage: cool [14:13] hello, i tried to install mpd with pulse-audio like described here http://mpd.wikia.com/wiki/PulseAudio on my ubuntu 14.10-server, but i stuck on this (http://mpd.wikia.com/wiki/PulseAudio#Adjusting_Volume_.2F_Sinks_.2F_etc.) step, i dont get it to connect to pulse using pacmd, it always says "No PulseAudio daemon running, or not running as session daemon" but i can see a pulseaudio process with "ps x" [14:17] LarsN: yes, hi [14:29] jamespage, looks like python-rtslib-fb needs to get into main for cinder, and python-wsgiref needs to be synced [14:30] coreycb, nack on rtslib [14:30] coreycb, just patch it out - see zuls pastebin above [14:31] jamespage, ok [14:38] coreycb, zul: can I get a +1 for including rabbitmq-server in the kilo CA please - 3.3.5 introduced better default queue failover [14:39] jamespage, +1 [14:39] jamespage, should I request a sync of python-wsgiref? [14:40] coreycb, I can't see that in debian or ubuntu? [14:40] any simplestreams folks around? [14:41] how does simplestreams get the latest version? is it lexical order? numeric order? [14:41] e.g. jq '.products."com.ubuntu.cloud.daily:server:14.10:amd64".versions |keys' com.ubuntu.cloud\:daily\:aws.json looks numeric, can I rely on that? === martins-afk is now known as martinst [14:54] jamespage, nevermind it's in libpython2.7-stdlib [15:06] hi - i'm just setting up server - and I need plenty of requests per second, started with ubuntu, and had nf_conntrack problem - fixed it, got better rps troughput - but I wonder is there linux distro which would be tuned for high permormance http serving out of the box? [15:06] or some links to sources about tuning ubuntu server? [15:23] jamespage, what do you think about getting a new vesion of pecan > 0.8.0 into vivid? (ceilometer) [15:25] coreycb, sounds like a good idea - I assume its required for ceilometer? [15:39] agend: not afaik. Are you server static content or an application. Typically you would tune your application to be faster. [15:59] jamespage, sorry, yes it's required by ceilometer [16:06] coreycb, ++ then [16:06] jrwren: content is dynamic, and not its not application problem - its written in go and it's pretty fast. I have to tune the os [16:09] agend: ah, so already profiled. Interesting. What issues are you seeing then? [16:34] coreycb, https://github.com/openstack/requirements/commit/70b10edb67ae72bf1f2436ee5abe6ae208ea4d2a [16:34] argh! [16:35] jrwren: like i said: nf_conntrack so far, got it better, now waiting for next surprise [16:35] jrwren: looks like i have to be ubuntu expert to get it done [16:36] jamespage, oh great [16:36] agend: i'm not sure what nf_conntrack is, but I've mostly run on cloudimg, which is a bit different than ubuntu-server === martinst is now known as martins-afk [16:36] jrwren: running on bare metal [16:38] coreycb, I'm seeing a nova test failure I think its related to that [16:40] jamespage, and 0.9.2 is the latest version it looks like [16:45] coreycb, yeah [16:45] coreycb, just thinking as to whether we should revert back to 0.9.1 [16:48] hi guys, once the server has booted up and services initialized, just before it goes to the console login prompt, the video mode seemingly changes… does anyone know where this is configured? as my kvm-switch doesn’t seem to support it. [16:49] jamespage, or temporarily revert the 0.9.2 change the caused it [16:50] i’m abou to try GRUB_GFXPAYLOAD_LINUX=keep [16:52] well that didn’t owrk === martins-afk is now known as martinst === martinst is now known as martins-afk [16:56] coreycb, actually it might be OK [16:57] coreycb, zul reverted the offending commit - I think I just has a bad test run locally - in ppa its fine [16:57] jamespage, ok [16:57] jamespage, https://code.launchpad.net/~corey.bryant/ubuntu/vivid/python-pecan/merge [17:00] coreycb, "Merge from Debian unstable?" [17:00] no? [17:00] coreycb, what are you merging? I'm confused [17:00] vivid has the same version as debian right now [17:01] coreycb, I was expecting 'New upstream release' [17:01] jamespage, ok.. yeah I guess we're forking in a way [17:01] coreycb, probably only for a bit - I've told zigo we're pushing kilo-1 updates this week [17:02] so he should have an eye out === markthomas|away is now known as markthomas [17:02] coreycb, merge == pull in changes from debian, re-apply ubuntu delta [17:02] coreycb, new upstream release == just ubuntu [17:02] coreycb, sync - just sync over from debian -> ubuntu [17:02] (drop any changes) [17:03] jamespage, yeah, makes sense [17:04] jamespage, ok pushed a new version [17:06] coreycb, for future reference, you don't need to say 'updated maintainer' - that's impled for a 0ubuntu1 [17:06] jamespage, ok [17:06] anyone? [17:07] coreycb, I'm this close to having nova ready I think [17:07] coreycb, unfortunately my power rate sync patch breaks tests now [17:07] * jamespage sighs [17:07] jamespage, :/ [17:08] coreycb, if https://launchpad.net/~james-page/+archive/ubuntu/vivid/+build/6661407 completes +1 [17:08] coreycb, swift is done btw [17:08] but that's all from my list [17:09] jamespage, ok. I'm a bit behind. ceilometer is just waiting on python-pecan and should be all set once we get the new version. cinder has a few more failing tests that I need to figure out. [17:09] jamespage, that's all I've made it to so far [17:11] coreycb, pecan uploaded - thanks! [17:11] jamespage, thank you! [17:13] after booting, just before the login prompt the screen resolution changes to something i don’t support… any ideas? (Ubuntu Server 14.04) [17:17] coreycb, nova - done [17:17] coreycb, and with that I'm off [17:18] jamespage, alright, thanks [17:58] huh, didn't know this channel existed. neat [17:59] repeating my other question then - is anyone aware if it is possible to control the installer dialog when doing a preseed/kickstart? === MeltedDed is now known as MeltedLux === martins-afk is now known as martinst === martinst is now known as martins-afk [19:06] Karunamon: you can suppress most dialogs by providing a pre-determined answer === Elion46 is now known as Elion === ggherdov`_ is now known as ggherdov` === markthomas is now known as markthomas|away === PaulePan1er is now known as PaulePanter [20:32] * mait is getting "AuthType Basic configured without corresponding module" from apache2 in 14.04.1 LTS [20:32] after upgrading from precise [20:42] they upgraded apache 2.2 to 2.4 so there are some changes. site config files need to end in .conf for example. [20:43] Got past some of those initial issues... === markthomas|away is now known as markthomas [20:44] I was thinking others might have seen this. My config looks very close to the example given for Apache 2.4 by apache.org [20:50] mait, enter apache2ctl -M [20:50] is auth_basic_module listed there? [20:52] Yes. Actually, the simple case works. My troubles are with AuthnProviderAlias [20:55] https://stackoverflow.com/questions/21385818/apache-authtype-not-set-internal-500-error/21386863#21386863 [20:56] mait, does that help? [20:59] Yes, that was part of what I've already gotten past. [21:00] could you paste your config file then? [21:00] Thanks, but I think there's something about 2.4 that doesn't work like 2.2's mod_authn_alias did. [21:01] mait, indeed. "The functionality provided by mod_authn_alias in previous versions (i.e., the AuthnProviderAlias directive) has been moved into mod_authn_core. " [21:01] https://httpd.apache.org/docs/2.4/upgrading.html [21:02] I'm trying to do the first example in http://httpd.apache.org/docs/2.4/mod/mod_authn_core.html [21:12] Hello all .. I'm not totally new at running my webserver, but I did a clean install of 14.04 (01) and now when I set up my sites they all point right back to the default Apache page . did something change? [21:14] NineTeen67Comet, what does ls /etc/apache2/sites-enabled/ return? [21:15] 000-default.conf and www.openlug.com.conf (my site) .. I've got more to add, but the first one was broken so I quit to seek knowledge :) .. [21:18] if you wanna remove 000-default.conf you can do so with "a2dissite 000-default" [21:20] Yeah .. headed in the right direction .. thank you much Meerkat . [21:20] then you need to enter "service apache2 reload" to load the new configuration. [21:20] Roger .. buttoned that one up .. I'm usually pretty good with the Ubuntu server (Been running it since I left Gentoo in 06 or so) .. So much thanks, I've been kicking myself all day .. [21:34] * mait may be hitting https://issues.apache.org/bugzilla/show_bug.cgi?id=55622 using apache2 2.4.7-1ubuntu4 [21:34] mait: Error: Could not parse XML returned by issues.apache.org: HTTP Error 404: Not Found [21:35] https://issues.apache.org/bugzilla/show_bug.cgi?id=55622 [21:35] mait: Error: Could not parse XML returned by issues.apache.org: HTTP Error 404: Not Found [21:36] Hmm. my broswer like it at least :( [21:42] it works for me too. [21:56] hi all, an apt-get nfs-kernel-server update on ubuntu 12.04 seems to have taken our production NFS share down :( [21:56] it's like read-only [21:57] Did you restart the nfs server after updating? [21:57] I just did that a few mins ago, did I do it right? I did sudo serviec nfs-kernel-server restart [21:58] logs on the server show [21:58] http://pastebin.com/bntfyJKJ [21:58] dmesg on the client machines show: [23366474.713247] NFS: state manager: check lease failed on NFSv4 server xyz.com with error 13 [21:59] Did you remount the share? [22:00] bekks: yeah, first I did "sudo umount -lf /mnt/nfs", then:: [22:00] sudo mount.nfs -o nosuid,nodev,noatime,nodiratime,noacl,nocto,rsize=32768,wsize=32768 xyz.com:/mnt/disk/nfs /mnt/nfs/ [22:01] it remounts but it's read-only... I get permission denied if I try to create or modify files [22:02] Whats the output of "mount | grep nfs" on the client? [22:03] nfs.xyz.com:/mnt/disk/nfs on /mnt/nfs type nfs (rw,nosuid,nodev,noatime,nodiratime,noacl,nocto,rsize=32768,wsize=32768,vers=4,addr=2a01:4f8:150:1309::2,clientaddr=2a01:4f8:150:30a8::2) [22:04] oh, you are using ipv6 [22:07] still getting permission denied all the time, it's weird [22:11] OK I turned on access logging (rpcdebug -m nfs -s proc) and I see: kernel: [3201308.065629] NFS reply getattr: 0 [22:35] OK so [22:35] anonuid=0 and anongid=0 saved us [22:35] the rpc idmapping thing is apparently not working right all of a sudden [22:35] maybe it is some ipv6 vs. ipv4 issue === bilde2910 is now known as bilde2910|away