[00:28] <adonaros> Hello team/group.. i seem to be having a bit of trouble with postfix and sieve. i think its simple. anyone want to take a stab at it?
[01:12] <patdk-lap> adonaros, postfix and sieve don't even work together
[05:34] <adonaros> patdk-lap: sorry
[05:34] <adonaros> dovecot **.
[06:05] <antares231> hello everyone. im complete beginner about linux server. where do i start to learn? sorry for my english :-[
[08:36] <frickler> jamespage: regarding https://bugs.launchpad.net/tripleo/+bug/1632538 once you have the new python-rfc3986 package available, I think you should also respin the python-nova package to explicitly require the new version. otherwise nova installations might still be broken unless people do package upgrades for everything
[08:37] <frickler> running our deployment with python-rfc3986==0.2.2 works fine now
[09:11] <rbasak> nacc: looks like there are still three unverified bacula SRU bugs. Please could you take a look?
[09:40] <jonah> Hi can anyone please help as I can't seem to start mysql this morning, I get this error: ExecStart=/usr/sbin/mysqld (code=exited, status=1/FAILURE)
[09:49] <Lartza> There should be an error
[09:50] <sarnold> jonah: check journalctl for log messages?
[09:52] <jonah> sarnold: getting this:
[09:52] <jonah> /usr/bin/mysql -v
[09:52] <jonah> ERROR 2002 (HY000): Can't connect to local MySQL server through socket '/var/run/mysqld/mysqld.sock' (2)
[09:53] <Lartza> ... journalctl -u mysql
[09:56] <sarnold> how about mysqld?
[09:57] <Lartza> sarnold, Afaik some genius named the service just "mysql"
[09:57] <Lartza> Not mariadb, not mysqld, mysql...
[09:58] <sarnold> ah
[11:14] <RoyK> Lartza: as you can see in the source code, the name of the internal bits hasn't been changed https://github.com/MariaDB/server
[11:15] <Lartza> RoyK, So?
[11:15] <RoyK> the command is still mysql, the daemon is mysqld etc
[11:15] <Lartza> Yet the systemd unit to start the daemon is mysql.service
[11:16] <RoyK> makes sense
[11:16] <Lartza> Not really
[11:16] <RoyK> anyway - feel free to file a bug
[11:17] <Lartza> mariadb ships a mariadb.service and iirc mysql-server had mysqld.service so mysql.service is just...
[11:17] <Lartza> I don't see how that makes sense
[11:22] <RoyK> Lartza: looks like you're right - I'm not sure which version of mariadb ubuntu has, though
[11:23] <Lartza> xenial seems to have 10.0.27
[11:31] <RoyK> Lartza: just checked in the mariadb 10.0 source - there it's called mysql.service
[11:32] <Lartza> Hmm :/
[11:32] <RoyK> ./dist/Ubuntu/mariadb-server-10.0.postinst:  deb-systemd-helper unmask mysql.service > /dev/null
[11:32] <RoyK> that's in server/debian
[11:36] <RoyK> https://github.com/MariaDB/server/blob/10.0/debian/dist/Debian/mariadb-server-10.0.postinst
[11:37] <sarnold> maybe check sources.debian.net instead, who knows how close 'upstream' mariadb is to our packaging
[12:10] <dlloyd> anyone else seeing SERVFAIL when resolving us.archive.ubuntu.com ?
[12:11] <sarnold> dlloyd: known issue, our dns providers are under ddos attack
[12:11] <dlloyd> ah ok
[12:11] <dlloyd> thanks
[12:44] <Pjusur> If I have two interfaces on the same subnett, but I want to enforce that all outgoing traffic goes from a specific interface, how do I accomplish this? IPtables?
[12:48] <sarnold> Pjusur: first guess, the 'ip' utility probably has a way to express this; maybe via setting one link up and another link down, or via routes, or perhaps you have to get familiar with http://lartc.org/howto/ to figure it out...
[12:54] <lordievader> Setting the metrics should do that, right?
[12:55] <sarnold> I'd hope so
[12:56] <sarnold> ahh, that's in ip-route. funny, I didn't expect that
[13:08] <frickler> coreycb: plugin.py(16): from openstackclient.common import utils
[13:25] <RoyK> hm - this canonical livepatch - how does that work with automatic updates? will the latter still upgrade kernels and ask for reboots? I can't find any docs on this
[13:28] <sarnold> RoyK: it should; some things can't be live-patched, and not all CVEs will be livepatched
[13:32] <coreycb> frickler, I'm not sure which client that plugin.py is coming from but I think probably one of your client packages is not upgraded to newton
[13:32] <RoyK> sarnold: ok, guess I'll just wait and see - never tried that sort of thing before
[13:32] <coreycb> frickler, http://paste.ubuntu.com/23359293/
[13:34] <joelio> RoyK: I doubt it needs a reboot, it'll be using kpatch/kexec one would assume. They'll be brief downtime, but not a wait for a full POST
[13:41] <sarnold> joelio: no, the live patching is using the in-kernel live patching mechanism. It's instant.
[13:42] <sarnold> joelio: there's some safeguards around making sure that processes currently executing in the kernel don't see the changes until after they return to userspace, but it should be visible to all processes on their next time entering the kernel
[13:43] <sarnold> joelio: this kind of live system patching of course won't work for everything; there's some things that can only be fixed by tearing the whole thing down via a reboot. So it's still important to reboot into new kernels occasoinally
[13:44] <beisner> hi coreycb, dosaboy - promoted python-oslo.service 0.9.0-1~cloud1 to liberty-updates in the cloud archiive re: https://bugs.launchpad.net/oslo.service/+bug/1524907
[13:44] <joelio> sarnold: interesting, I'd just assumed it was leveraging kpatch/kexec.. will do some digging into this
[13:45] <sarnold> joelio: http://chrisarges.net/2015/09/21/livepatch-on-ubuntu.html
[13:45] <joelio> sarnold: thanks
[13:46] <joelio> sarnold: that mentions kpatch?
[13:47] <sarnold> joelio: yeah, sorry, I focussed entirely on 'kexec' in your original statement, heh
[13:49] <joelio> sarnold: no worries :)
[13:52] <coreycb> thanks beisner
[14:07] <powersj> rbasak: during triage last night came across lp# 1483093 which is marked incomplete for >60 days, any idea why it showed up?
[14:11] <rbasak> powersj: not sure. https://bugs.launchpad.net/ubuntu/+source/squid-deb-proxy/+bug/1483093/+activity doesn't suggest anything. Did I introduce a bug in the script?
[14:12] <powersj> rbasak: I wondered that, however if you look at all the bugs page for squid-deb-proxy it shows up there as well
[14:13] <coreycb> beisner, can you do a full sweep of liberty-proposed -> liberty-updates for bug 1619452?
[14:16] <rbasak> powersj: for me its ordering looks correct in https://bugs.launchpad.net/ubuntu/+source/squid-deb-proxy/+bugs?orderby=-date_last_updated&start=0
[14:18] <powersj> rbasak: I agree, I guess I did not expect to see that bug at all given it should have expired, but I just realized it is assigned to someone so it won't expire
[14:20] <rbasak> Ah
[14:20] <rbasak> I thought the problem was that my script showed it up when not requesting a 2015 date range? Or did that not happen?
[14:20] <rbasak> Or is the script supposed to show it up?
[14:20] <powersj> no the query that I ran was for 2015, so it correctly showed up
[14:20] <rbasak> Ah
[14:20] <rbasak> Shall we unassign it?
[14:21] <powersj> yeah
[14:21] <rbasak> Done
[14:21] <powersj> cool :) thx
[14:21] <rbasak> (I didn't feel that needed an explanation as the bug is Incomplete and clear on what needs to happen next anyway)
[14:21] <powersj> agreed
[14:36] <rabbitdew> I just added a hard disk to my ubuntu VM to expand the drive on /var. I'm using LVM but I forgot to use fdisk to write a partition table to the new disk and format it LVM before adding it as a pv, growing the vg, extending the lv, and expanding the filesystem
[14:37] <rabbitdew> I don't think there's anyway that filesystem has started to fill up onto the unpartitioned space. But partprobe now throws 'Error: Can't have a partition outside the disk!' I'm really concerned it's going to bork out. Any ideas on how I can revert this?
[14:38] <rabbitdew> I'm using ext4. Can only expand a mounted fs
[14:51] <rbasak> What did you add as the pv? The whole disk then and not a partition?
[14:51] <rabbitdew> yes
[14:51] <rbasak> AFAIK, that's an acceptable and stable situation. Partitioning tools won't be able to understand the disk, that's all.
[14:51] <rabbitdew> THe whole disk (/dev/sdb)
[14:52] <rbasak> I'm not aware of anything that will let you change that without severe hackery (except for removing the data and dropping the pv, etc)
[14:52] <RoyK> rabbitdew: so pvcreate /dev/sdb ; vgextend somevg /dev/sdb ; lvextend -L something /path/to/my_lv?
[14:53] <rabbitdew> RoyK: yep, that's exactly what I did
[14:53] <RoyK> and then just resize2fs /dev/whatever ?
[14:54] <rabbitdew> Yeah
[14:54] <RoyK> no arguments to resize2fs?
[14:54] <rbasak> I see that pvresize exists.
[14:54] <rabbitdew> resize2fs /dev/logicalvolume
[14:55] <RoyK> rabbitdew: can you pastebin it all, please, including messages?
[14:55] <rbasak> So perhaps you could use that to shrink the PV a little, then use some tool to move the data up the disk, then create a partition table that fits where the data is. I don't know to what extent LVM will accept that, and that does qualify as severe hackery in my book.
[14:55] <RoyK> pvmove
[14:55] <rbasak> Oh, neat.
[14:55] <rbasak> But still, creating the partition table after the data is there is pretty hacky!
[14:56] <RoyK> just don't do it
[14:56] <rbasak> Yeah I agree
[14:56] <rbasak> Not worth it going wrong.
[14:56] <RoyK> rabbitdew: how big was the lv before you extended it?
[14:56] <rbasak> (or worse, leaving a latent problem for it to explode in a few years' time)
[14:58] <RoyK> rabbitdew: first of all - do you have a good backup?
[15:00] <rabbitdew> =) No I sure don't
[15:00] <rabbitdew> I had just created the vm yesterday as a logstash server
[15:00] <rabbitdew> http://pastebin.com/XmDefqHn
[15:00] <RoyK> rabbitdew: then - first of all - make one - there are cloud services out there that work well, crashplan, for instance
[15:01] <RoyK> rabbitdew: what's the output of vgs?
[15:04] <beisner> hi coreycb - yep, on that liberty sru / point release pkgs now.
[15:04] <rabbitdew> RoyK: I'm pretty sure it was 60GB before I resized it. Here's vgs and vgdisplay http://pastebin.com/FE2gsZmP
[15:04] <RoyK> rabbitdew: what about pvs and lvs?
[15:06] <rabbitdew> http://pastebin.com/nZY0t2KR
[15:06] <RoyK> rabbitdew: next time, if you want to use the whole disk for expansion, just do lvextend -l +100%FREE
[15:07] <rabbitdew> RoyK: Ah, that's what I had wanted to do. I'll do that in the future.
[15:07] <RoyK> you can do it now as well
[15:07] <RoyK> anyway - what did resize2fs say?
[15:12] <RoyK> rabbitdew: partprobe now throws 'Error: Can't have a partition outside the disk!' <-- I've seen similar messages when not using partitions
[15:13] <rabbitdew> I closed the terminal I had it open, so I don't have its output. But it didn't throw any errors
[15:13] <RoyK> parts in Linux seem to default to the thought of that everything must be on a partition
[15:14] <rabbitdew> It really might be cool, but I should put a good backup plan into place
[15:14] <RoyK> never had issues with it, though, apart from those messages
[15:14] <rabbitdew> I'll know later today
[15:15] <rabbitdew> I should have more than 60G of data written to it by later this afternoon
[15:15] <RoyK> anyway, please give backups a thought
[15:15] <RoyK> disks fail, it's in their nature, like everything else
[15:15] <rabbitdew> Thanks a lot for helping, also thanks rbasak
[15:16] <rabbitdew> Yeah, I hear you. I'm pretty good about backing up my workstation/laptop and everything else. Because of politics I had to fight to get a logstash server in the first place
[15:16] <rabbitdew> Disk space is cheap
[15:17] <RoyK> rabbitdew: if this is work related, I hope you have a good, centralised backup solution ;)
[15:22] <nacc> rbasak: ack, will do that to day
[15:22] <nacc> *today
[15:29] <caravaggio971> hi there
[15:29] <caravaggio971> can anyone help me in restoring a backup i made with rsnapshot?
[15:30] <theGoat> so updated one of my ubuntu instances to 16.04, i have added service to systemd for freeradius.  now when i do a service freeradius start, it just hangs...when i control c out of and do a ps it's running....not sure what's hanging it.  how can i debug it?
[15:30] <caravaggio971> #ubuntu-it
[16:02] <nacc> jgrimm: fyi, i think the imports are all updated in usd-import-team
[16:17] <jgrimm> nacc, cool, i'll take a look today
[16:18] <nacc> jgrimm: thanks
[16:37] <misterpink> anyone successfully setup dropbox on ubuntu server?  tried 100 different ways and always get a permissions error.  used an older version of dropbox that worked however it is no longer supported
[16:54] <misterpink> "unset DISPLAY" solved the permissions error issue
[17:00] <nacc> rbasak: i think i got them all, can you confirm?
[17:54] <ivoks> nacc: rbasak https://bugs.launchpad.net/ubuntu/+source/bacula/+bug/1553563 ; fix confirmed
[18:07] <stoned> Hello, is it advisable to upgrade from 14 LTS to 16 LTS?
[18:08] <RoyK> stoned: do you need to?
[18:08] <stoned> Or should I spin up a new server for 16 and go from there
[18:08] <RoyK> stoned: if you need the new stuff in 16.04, just do-release-upgrade the 14.04 one - should work fine
[18:09] <stoned> Well, kind of.
[18:09] <RoyK> stoned: if you don't need the new stuff, the old one will work fine and it'll be supported until march 2019
[18:09] <stoned> Or actually, maybe not, but... to save money
[18:09] <stoned> I need to run a website, try some web software that is only available via snaps
[18:09] <stoned> Rocketchat server for chat.stoned.io
[18:09] <RoyK> setup a vm, then
[18:09] <stoned> yeah, another vm is $20 bucks.
[18:09] <stoned> :)
[18:10]  * RoyK mostly runs his stuff on his own hardware
[18:12] <stoned> you must have good b/w
[18:12] <stoned> and good h/w
[18:12] <stoned> my desktop  is this
[18:12] <stoned> OS: Linux 3.16.0-4-amd64/x86_64-Distro: Debian 8.6-CPU: 4 x Intel Core2 Quad (3775.073 MHz)-Processes: 670-Uptime: 1d 4h 4m-Users: 6-Load Average: 0.21-Memory Usage: 3711.03MB/8002.49MB (-15.64%)-Disk Usage: 3392.89GB/3734.40GB (90.86%)
[18:12] <stoned> that's all I go.
[18:12] <stoned> got*
[18:13] <stoned> Who is a good ubuntu server vps provider that cheap/affordable/reliable etc. ? or is that not a good quetion here?
[18:14] <RoyK> stoned: should be sufficient to setup a few VMs
[18:15] <RoyK> a small webserver won't need more than half a gigabyte of RAM anyway, and CPU doesn't seem to be your bottleneck
[18:18] <stoned> Well, I'm on really crappy bandwidth.
[18:18] <RoyK> how much?
[18:19] <stoned> it's more about reliability, the lines here suck
[18:19] <RoyK> I see
[18:19] <stoned> it' goes out like every couple of hours if not less
[18:19] <RoyK> then - what is it you need for that new webserver that isn't in 14.04?
[18:20] <stoned> I think this thing called snaps, the rocketchat server is only available from it
[18:20] <RoyK> stoned: heh - I switched to a new ISP some five years back and the worst downtime I've had since then was an hour or two when one of their servers went down (plus perhaps an annual hang on the router)
[18:20] <stoned> I don't even really  need to upgrade, I prefer older/stable/secure software
[18:20] <RoyK> right - snaps isn't in 14.04 and probably won't be
[18:21] <stoned> <- comcast customer
[18:21] <stoned> woe is me.
[18:21] <stoned> here is my currnet problem
[18:22] <RoyK> stoned: what sort of vm do you currently have? openvz or kvm?
[18:23] <stoned> https://paste.debian.net/plainh/f8575843
[18:23] <stoned> This is my current issue
[18:23] <stoned> I'm on  Rackspace ATM and with ubuntu 14
[18:23] <stoned> I just wanted to play around with some chat backends
[18:23] <stoned> I guess, I should spin up another server for playing around and leave this alone?
[18:24] <RoyK> stoned: check crowncloud.net - doesn't cost a lot
[18:24] <RoyK> I have a vm there and a friend has another (which I'm mostly adminstering) - works well
[18:25] <stoned> yeah, I wanted to setup owncloud, my own vpn, and all the stuff. I kknow how to, I just need to find a good stable server. I'd stick w/ RS, but I don' tknow amazon AWS. I wonder if I should spend some money there, and learn how to use AWS and add that to my skillset.
[18:25] <stoned> So many options.
[18:25] <RoyK> crowncloud != owncloud ;)
[18:25] <stoned> But that's not really a question here
[18:26] <stoned> yes, I know :)
[18:26] <stoned> It just reminded me of it. similar word
[18:26] <RoyK> stoned: how much do you pay for your server today?
[18:26] <stoned> I don't
[18:26] <RoyK> ok
[18:26] <stoned> https://paste.debian.net/plainh/f8575843 read this
[18:26] <stoned> it will explai my issue
[18:27] <RoyK> if you can handle $4/month for a 1GB RAM thing from crowncloud, go for it
[18:27] <stoned> Sure, I can handle it.
[18:27] <stoned> Here's another thing.
[18:28] <stoned> I want to setup a server instance using a config managemenet tool
[18:28] <RoyK> but I'd advice against imaging the server - better setup a new one from scratch
[18:28] <RoyK> what sorts?
[18:28] <stoned> So I can use just that, create a config, use a tool to deploy the same server image locally in vm, and in cloud as a server instance
[18:28] <stoned> That way I can keep one config ile, and deploy servers anywhere
[18:28] <stoned> same config, same seruciry, packates, hardening, etc. etc.
[18:28] <RoyK> what sort of config management?
[18:28] <stoned> Just that on local, i can change a few php settings for dev.
[18:29] <stoned> puphhete
[18:29] <RoyK> puppet
[18:29] <RoyK> ok
[18:29] <stoned> phupphet vagrant has a rackspcae plugin
[18:29] <stoned> you can deploy ubuntu server imaes yuo create from this config file to local vm and to rackspace
[18:29] <stoned> so maybe I'll have to stick with RS for now.
[18:29] <stoned> It's $20/mo, more expensive, but whatever. It's more convenient.
[18:30] <RoyK> again, is this a kvm vm or openvz?
[18:31] <RoyK> if it's a kvm vm, you should be able to deploy an openvz container on top
[18:32] <stoned> no clue
[18:32] <stoned> I use vbox
[18:32] <RoyK> run virt-what
[18:33] <stoned> vagratn/vbox locally
[18:33] <RoyK> on the server
[18:33] <stoned> on the server, no clue
[18:33] <stoned> it's rackspace
[18:33] <stoned> Xen?
[18:33] <RoyK> virt-what will tell
[18:33] <stoned> oh
[18:33] <stoned> hang on
[18:33] <stoned> The program 'virt-what' is currently not installed. You can install it by typing:
[18:33] <stoned> apt-get install virt-what
[18:33] <stoned> hm.
[18:33] <RoyK> then install it :P
[18:34] <stoned> xen
[18:34] <stoned> xen-domU
[18:34] <stoned> yup, RS uses Xen
[18:35] <RoyK> IIRC they're rolling out kvm now, but they probably have a *lot* of old xen left
[18:35] <RoyK> anyway - with a machine on xen, you should be able to spawn an openvz container on top of that
[18:37] <stoned> hm.
[18:37] <stoned> interesting.
[18:37] <stoned> it's only got 512mb ram
[18:37] <RoyK> that's not a lot
[18:37] <stoned> yeah, cheapest one.
[18:37] <stoned> Maybe I should move away from Rackspace.
[18:37] <RoyK> so probably better get something off crowncloud :P
[18:37] <stoned> I'm checking out crowncloud
[18:37] <stoned> https://lowendbox.com/blog/crowncloud-30year-3gb-openvz-vps-and-more-in-los-angeles/
[18:38] <RoyK> I have a VM there running Zimbra with 5GB RAM and 140GB disk, costs me $16 a month
[18:38] <RoyK> https://crowncloud.net/openvz.php
[18:38] <stoned> Hm.
[18:39] <stoned> I can pay them $30 for 1 year of 2gb server?
[18:39] <stoned> cheap is usually not that great.
[18:39] <RoyK> well, I've used them for a while now, and it works well
[18:40] <RoyK> and zimbra isn't really a light-weight system
[18:40] <RoyK> perhaps I should have migrated to something else - but it works - so why bother :P
[18:45] <RoyK> stoned: if you have questions to crowncloud, the fastest way is usually to ask them on #CrownCloud here on freenode
[19:10] <nacc> ivoks: thanks!
[20:28] <dcampano> curious if people running 16.04 server use the standard timesyncd or if they install ntp?
[20:30] <docmur> I have a server running 16.04.1 LTS, I install virt-manager and libvirt, when I ssh to the server using -Y -2 I get X11 forwarding request failed on channel 0 and when I try to run virt-manager I get Failed to connect to Mir: Failed to connect to server socket: No such file or directory
[20:34] <sarnold> docmur: try -X instead; debian fiddled with the meanings of -X and -Y
[20:34] <sarnold> docmur: actually, try DISPLAY=:10 virt-manager  before trying -X
[20:37] <docmur> using ssh user@host DISPLAY=:10 virt-manager did nothing and using -X is also not working :(