[07:07] rharper: powersj: my qemu testing found the first bug we didn't have on our radar yet [07:07] so while still uncomplete to an ashaming amoutn already worth something [07:08] I found an upstream qemu bug and a fix, added Ubuntu(qemu) to it and subscribed the server Team [07:08] rbasak: fyi ^^ [07:57] cpaelzer: good job! [08:56] coreycb, figured out the barbican test failures - we need newer positional [08:56] doing that now [11:44] jamespage, hello, when you get a chance heat 1:5.0.1-0ubuntu3.1~cloud1 is ready to promote to liberty-proposed [12:11] jamespage, any news on barbican? [12:11] ddellav, how's openstack-trove going? [12:12] coreycb ah forgot about that one, i think it's done, let me double check === freyes__ is now known as freyes [13:10] ddellav, jamespage: I took a pass through newton pkg tip failures and rebuilding those now (non-barbican/trove) [13:26] coreycb, barbican should build now I think [13:26] * jamespage checks [13:26] jamespage, cool [13:27] coreycb, it was https://launchpad.net/ubuntu/+source/python-positional/1.1.1-0ubuntu1 [13:27] jamespage, ah just needed a new version? [13:28] coreycb, yeah I think so [13:28] jamespage, awesome. that wasn't obvious from what I recall. [13:29] anyone has used isc-dhcp-server as a dhcp server in a multivlan environment, it's leasing very very slow [13:29] can anyone help? [13:38] coreycb, ok barbican is happy again [13:39] coreycb, I'll do the b3 upload now [13:39] jamespage, \o/ [14:13] coreycb, ddellav: any ideas on the trove failure? [14:13] that's our last outstanding I think [14:16] jamespage, ddellav: just looking at tox.ini for trove and they run ostestr --serial [14:20] coreycb, might be it [14:20] worth a try [14:24] coreycb, http://autopkgtest.ubuntu.com/packages/h/horizon/yakkety/amd64 [14:24] hmm [14:24] jamespage, I thought we got rid of openstack-dashboard-ubuntu-theme [14:25] #not sure [14:25] coreycb, you look at trove [14:25] I'll poke at that [14:27] jamespage, ok. that's odd because I was able to upgrade from b2 successfully. might be something to do with charm config vs pkg config. [14:33] coreycb, yeah I see that on a fresh install [14:34] jamespage, ok [14:35] jamespage, trove has a dep on xmltodict which is in universe [14:36] coreycb, as its in universe that's just fine :-) [14:36] jamespage, duh yeah :) [14:37] coreycb, trove built - good-oh [14:37] jamespage, yep! I'll get that uploaded. [14:39] coreycb trove b3 was waiting on openstackdocstheme, thats why i wasn't pushed. It's good to go now and building, ready for push: lp:~ddellav/ubuntu/+source/openstack-trove [14:41] ddellav, jamespage: trove's uploaded [14:46] grumble.. I wish ubuntu cloud team would get their act together with ubuntu/xenial64 images (specifically for vagrant). Just hit by a bug which means I can't create more than one ubuntu/xenial64 box as it names them all the same! trusty is fine and using chef's is fine... just ubuntu can't... ubuntu. [14:48] everything gets the name... [14:48] ==> gateway1: Setting the name of the VM: ubuntu-xenial-16.04-cloudimg [14:48] obviously that fails hard when using multivm vagrant :) [14:49] exact same config, but using trusty... [14:49] ==> gateway1: Setting the name of the VM: devstack-vm_gateway1_1473259350488_72222 [14:49] lovely :) [14:49] want me to raise a bug? [14:55] beisner, jamespage: the following are ready for promotion when you get a chance, please: http://paste.ubuntu.com/23146098/ [14:58] joelio: The fixes for that have landed today, and we should hopefully be rolling out a fixed image in the next day or so. [14:58] Odd_Bloke: no worries, I appreciate this is a group effort [14:59] joelio: :) [14:59] can switch to trusty for this dev tranche, but will roll out Xenial when we're doing properly [14:59] thankfully it's nothing too taxing that required xenial or somesuch :) [15:00] Odd_Bloke: aware if that new image fixes vboxsf or not ooi? [15:20] smb, hallyn: is there any regression bucket you run against qemu when making changes? or is DEP8 sufficient? [15:23] jgrimm, I did not do substantial changes to qemu, so dep8 was usually enough for me. Serge might have been running more [15:24] smb, cool, yeah fix is mostly isolated to a specific use case on ppc64 architecture, but i'd like to be safe as possible [15:30] jgrimm: lp:qa-regression-testing [15:30] * jgrimm looks. thank you hallyn [15:30] run at least test-qemu.py, probably also test-libvirt.py [15:30] got it. thanks! [15:31] you'll want to do that in the cloud probably to minimize your own bandwidth usage :) unless you're one ofthose lucky googlefiber ppl [15:31] np - \o [15:31] :) [15:31] btw, for changes to qemu in yakkety i still think it would be best to keep them in sync in the ubuntu-dev branch of the debian packaging tree [15:31] and, i do mean to keep being helpful with that pkg, but things are a bit manic for the next month or so probably [15:32] but shout if you need anything, and i fyou have a pkg you want me to test, that doesn't actualy take me a lot of time. [15:32] hallyn, i deeply appreciate your continued help! [15:32] (my container hosting vm is kind of setup to make qemu testing easy :) [15:32] np - \o [15:32] hallyn, i have a ppa here if you'd like: ppa:jgrimm/qemu-1541902 [15:33] which release is that for? [15:33] hallyn, yakkety for this: https://bugs.launchpad.net/ubuntu/+source/qemu/+bug/1541902 [15:33] Error: Could not gather data from Launchpad for bug #1541902 (https://launchpad.net/bugs/1541902). The error has been logged [15:33] ok [15:34] hallyn, ibm offered to help with the unique hardware / use case, so that's covered === RoyK is now known as Guest12777 === RoyK^ is now known as RoyK [15:56] coreycb, ok horizon fixes uploaded === petevg_afk is now known as petevg [16:00] jamespage, \o/ looks like that one required some good eye sight. a dash changed to an underscore. [16:53] Odd_Bloke, hey - are you aware of any issues with out yakkety cloud images? [16:53] I get this: [16:53] E: Malformed entry 11 in list file /etc/apt/sources.list (URI parse) [16:53] when trying to update packages [16:53] appears to be double space between deb[-src] yakkety [16:54] actually that foorbar [16:54] http://paste.ubuntu.com/23146736/ [16:55] * rbasak wonders if that's cloud-init [16:55] As there has been some work in cloud-init in that area recently [16:55] cpaelzer, smoser: ^? [16:56] :-( [16:56] jamespage, where do you see this ? [16:58] smoser, that was in a yakkety environment I just tried to deploy for openstack [16:58] http://paste.ubuntu.com/23146752/ [16:58] so it does not reproduce ^ (lxc) [16:59] ii cloud-init 0.7.7-22-g763f403-0ubuntu1 all Init scripts for cloud instances [16:59] can you get me a cloud-init.log file where it happened ? [16:59] smoser, http://paste.ubuntu.com/23146756/ [17:00] smoser, I got that all all 15 instances [17:01] Sep 7 16:19:28 ubuntu [CLOUDINIT] cc_apt_configure.py[DEBUG]: got primary mirror: [17:01] hmm [17:01] Sep 7 16:19:27 ubuntu [CLOUDINIT] cc_apt_configure.py[DEBUG]: apt config: convert V2 to V3 format for keys 'apt_mirror' [17:01] Sep 7 16:19:27 ubuntu [CLOUDINIT] cc_apt_configure.py[DEBUG]: handling apt (module apt-configure) with apt config '{'primary': [{'uri': '', 'arches': ['default']}]}' [17:01] jamespage, did you provide some config ? [17:02] smoser, nope - its was juju instantiated [17:02] hm. [17:02] machine 0 is xenial - that came up ok [17:02] am i able to get in ? [17:03] smoser, yes see pm [17:17] If I run `conjure-up openstack` on the MAAS region controller, what is needed for MAAS to be an option in the initial menu (other than `localhost`)? [17:18] I remember it just being an option a month or so ago, but it doesn't seem to show up now (MAAS server is a fresh reinstall as of a few days ago) [17:33] What ports do I open in the firewall for a Postfix MTA? [17:33] Google just told me to read main.cf, which has zero references to ports [17:36] Oh hey, after some finagling, I got a stack overflow question for it: 25, 465, and 587 === alexisb is now known as alexisb-afk [17:46] do what? who ever answered you is kindof insane [17:46] 465 was NEVER an email port [17:46] patdk-wk: google disagrees, something about SMTP + SSL [17:46] 25 and 587 are the only two [17:47] I thought google was a search engine [17:47] patdk-wk: https://en.wikipedia.org/wiki/Simple_Mail_Transfer_Protocol [17:47] you point me to the RFC that says to use 465 [17:47] 465 NEVER made it into a *beta* version of the smtp standard [17:47] patdk-wk: do you also complain every time someone mentions 6667 is the IRC port [17:47] patdk-wk: because that was never in the RFC either [17:48] jerichowasahoax, that is completely different [17:48] no one moved to ssl+smtp [17:48] patdk-wk: it's different because you're already yelling at me about SMTP? [17:48] everyone continues to use port 25 [17:48] patdk-wk: because at this point i'm pretty sure you're attacking me for the sake of attacking me [17:48] irc branched out, to other ports, due to the 64k limit [17:50] http://blog.mailgun.com/25-465-587-what-port-should-i-use/ [17:51] that is actually a good write up [17:59] saying that it never was an email port is wrong, of course; it was registered with IANA and was used in the wild [18:00] but you shouldn't use it in any mailserver nowadays [18:00] registered and revoked, and never made it into any publications [18:00] jgrimm: tests passed [18:01] hallyn, \o/ thanks. ill run them too, just so i learn how to be self-sufficient [18:01] it's been revoked for almost 20years [18:02] right [18:05] if 465 winds up getting unused i'll just close the port again [18:05] "sudo ufw delete allow 465" is not that difficult :V [18:05] that isn't the point [18:05] the point is to not have users misconfiguring their mail software in the first place [18:06] postfix is default configuration except for I switched to maildir [18:06] to have them use 587, with login [18:06] a default configuration is NEVER recommended [18:06] and if your using ubuntu, you cannot get a postfix default config, you can get a debian config of postfix though [18:07] well i mean i obviously didn't leave the domain as example.com or anything [18:07] "default configuration" means "anything not mentioned in the guide i'm following" [18:07] which is this https://www.digitalocean.com/community/tutorials/how-to-install-and-configure-postfix-on-ubuntu-16-04 [18:07] I see the examples in 16.04 are a lot more sane though, that is nice === alexisb-afk is now known as alexisb [19:27] hi all :) i'm trying to increase mysqld's fd limit. i edited /etc/security/limits.conf and rebooted. if i open a shell with "su -l mysql -s /bin/sh" and evaluate "ulimit -Hn" or "ulimit -Sn" I see 65536. that's good. i edited /etc/mysql/mysql.conf.d/mysqld.conf with "open_files_limit=4096" but if i ask mysqld "select @@open_files_limit;" i still get 1024. is something somewhere else configuring limits for mysqld? [20:17] Hello, Am I in right channel I am not sure. But I have a problem. I have a server and I installed ubuntu server. I want to do virtualization. Actually I want to use container something like LXC to create more machine. . [20:17] But If I create machine with LXC so Can I install docker and kubernates to that? Do you have any idea or experience? [20:22] temhaa: do you want to do virtualization or containers? [20:24] nacc, Actually I want to virtualization with KVM but I wil try container(lXC) to open more machine. My question is If I use LXC, Can I install docker or kubernates to inside of the that container [20:26] temhaa: sorry, your first sentence is still confusing to me -- do you want KVM or LXC? If you don't want KVM, just don't mention it. For information on docker under LXD: https://www.stgraber.org/2016/04/13/lxd-2-0-docker-in-lxd-712/ and you might want to join #lxcontainers [20:28] nacc, choosing the virtualization method is another problem for me. But my question is If I use lxc can I install docker and kubernates to that lxc container. [20:28] nacc, I can not find more documentation on internet. I saw post what you said. But I am looking for kubernates [20:29] docker you can, i am not sure about kubernates .. [20:29] nacc, I asked in lxc and kubernates channels but I couldnt get any answer [20:31] temhaa: well, you said docker or kubernetes -- so you only want kubernetes, really? [20:34] nacc, I need docker but additionaly I want to install kubernates. They can not be at same container. [20:34] temhaa: https://github.com/kubernetes/kubernetes/issues/6862 ? [20:37] nacc, so we can not ~~ [20:38] temhaa: i'm not 100% myself, which lxc channel did you ask in? [20:39] stgraber: --^ can you chime in? [20:39] nacc, I asked to lxc channel called "lxc" and I asked to kubernates on slack [20:42] sorry, never used kubernetes myself. My understanding is that it currently relies on Docker containers and those do work (for the most part) inside LXD containers, but that's the extent of my knowledge there [20:43] stgraber: thanks [20:51] hi can anyone please help. I'm trying to tar up some files to move a website to a different server but whatever I do some of the files when untarred are empty [20:51] I can download them individually and upload them one by one but can't tar them up! [20:51] There are thousands of files so really need to tar them to move them... [20:51] If anyone can please help. I've tried zip, tar tar.bz2 etc, every time it is the same. [20:51] Most files do tar ok and extract ok, but some crucial site system files extract and just have 0 bytes and then cause erros on the site loading [20:55] jonah: can you give a concrete example of the command that you used, ls -l output of what you're trying to archive, tar tf output of the tarfile once it's created, tar xvf output, and ls -l of the directory structure after untarring? [22:32] should i set up LVM? it's an option on my guided partitioning method thing.. LVM, LV root and LV home-vg [22:34] nm went around it [22:36] I always use lvm [22:38] is it easy to set the internal IP to static so it doesn't get reassigned when there's a power outtage [22:45] skulltip: /etc/network/interfaces [22:45] skulltip: be sure to modify your network's dhcp server to not hand out that address to clients [22:50] internal ip? [22:55] on clean install of the server 16.04.1, it's stuck with first message: /dev/sda1: clean: xxxx / xxx files: 41081418344 blocks [22:55] is it hung? [22:55] i'm googling that fcsk is running? [22:56] that sure sounds like a message from fsck [22:56] do i need to wait for it to finish then [22:57] might be waiting for the network to come up? [22:58] skulltip: it wouldn't be a bad idea to wait, that will mark the date the scan completed and hopeflly push off the date to the next fsck, rather than jhust run it again on the next boot