[00:16] <arooni> about to upgrade ubuntu 14.04 => 16.04 server via command line; anything i need to keep in mind?
[00:16] <arooni> about to upgrade ubuntu 14.04 => 16.04 server via command line; anything i need to keep in minabout to upgrade ubuntu 14.04 => 16.04 server via command line; anything i need to keep in mind?d?
[00:16] <arooni> oops sorry for the repost
[00:16] <arooni> (twice :( )
[00:17] <sarnold> if you've got apache installed I think 16.04 changed the authentication and authorization stuff
[00:17] <sarnold> and php7 doesn't handle everything php5 did. or something like that.
[00:18] <rbasak> arooni: take a full system backup before you begin :)
[00:18] <arooni> is doing the upgrade not recommended
[00:18] <arooni> i.e. as opposed to transferring to a clean install
[00:19] <sarnold> no, upgrade should work, please file bugs if you run into trouble. but backups are always a good idea. :)
[00:20] <arooni> running a snapshot backup first
[00:29] <arooni> i wasnt really paying attention; was running sudo apt-get dist-upgrade wrong
[00:30] <sarnold> do-release-upgrade is the supported way to move from one release to the next
[00:34] <arooni> sarnold: ok so a bit of dist-upgrade already ran before i control c'd it
[00:34] <arooni> does that mean i really screwed up
[00:34] <sarnold> maybe
[00:35] <arooni> and need to restore from the backup before trying again
[00:35] <sarnold> did it start unpacking packages yet?
[00:35] <sarnold> or was it still downloading?
[00:35] <arooni> ummm i think it was setting up packages :(
[00:36] <arooni> got to this https://gist.github.com/5b17ced2bbdde3cae2cc806cdba9bae8
[00:36] <sarnold> *probably* if you set the apt sources back the way they were, and run do-release-upgrade, it'll sort out a path forward
[00:36] <sarnold> arooni: those numbers seem sane. might as well let it go :)
[00:37] <arooni> well i have a backup in case i kaboshed it
[00:37] <arooni> may as well see
[00:45] <sdeziel> arooni: if you have apache2, sites-enabled needs to contain files ending with .conf
[00:45] <rbasak> nacc_: do you want me to work on getting gpgv in the snap tomorrow? Or are you looking at it already?
[00:46] <arooni> sdeziel: i'm using nginx
[00:46] <arooni> if there are any gotchas there
[00:46] <sdeziel> arooni: not that I remember
[00:46] <sdeziel> arooni: only good things ahead (http2) :)
[00:47] <nacc_> rbasak: i can work on it now, and i'll send an mp off yours?
[00:47] <nacc_> it should end up passing ci as a means to check it
[00:49] <arooni> goodbye php 5.6
[00:49] <arooni> hello php 7
[00:50] <nacc_> arooni: 14.04 -> 16.04?
[00:50] <arooni> howd you guess lol
[00:51] <rbasak> nacc_: sure, thanks
[00:51] <nacc_> arooni: :)
[00:51] <arooni> should i keep the unattended upgrades config file i created
[00:51] <arooni> diff doesnt appear to have anything about a specificrelease
[00:56] <sdeziel> arooni: trusty ships with php 5.5.9, do you have any PPAs?
[00:57] <arooni> i think i did
[00:57] <arooni> probably   should have blown them away      prior to upgrade
[00:57] <nacc_> do-release-upgrade does automatically disable them, but you still need to purge them, iirc
[00:58] <arooni> well it booted up successfully as 16.0.4
[00:58] <arooni> 16.04.4
[00:59] <nacc_> arooni: well, that's not released yet, so hopefully not :)
[01:06] <sdeziel> nacc_: the base-files update was pushed already
[01:06] <nacc_> sdeziel: really?
[01:06] <nacc_> sdeziel: that's surprising, but i guess doesn't hurt
[01:06] <sdeziel> grep VERSION= /etc/os-release
[01:07] <sdeziel> all of my systems are reporting 16.04.4
[01:07] <arooni> so ;;; if there are things listed in my sources.list.d directory that reference trusty
[01:07] <nacc_> sdeziel: interesting, i thought that would be coordinated with the official release, but maybe that's only for the iso
[01:07] <arooni> do i need to delete/remove them
[01:07] <arooni> or will they not be used as its not the xenial release
[01:07] <sdeziel> nacc_: the ISO always seems to lag from a couple days
[01:08] <sdeziel> arooni: you need to delete/update them
[01:08] <nacc_> sdeziel: well, the iso isn't supposed to release until mar 1
[01:08] <sdeziel> arooni: well, only the files with names ending with .list
[01:09] <arooni> what are the other files there for
[01:09] <sdeziel> leftovers from the release upgrade mostly
[01:24] <nacc_> rbasak: i'm testing the trivial revert now, and then i'll need to narrow it done
[01:24] <nacc_> i should be able to work on it more tmrw
[03:17] <nacc_> rbasak: sigh, you're calling gpg too in the tests :)
[03:19] <nacc_> rbasak: i think we can snap all of gpg again, but not use it in our code
[03:19] <nacc_> beyond the tests
[03:19] <nacc_> i'll need to verify that's actually the case, though
[03:58] <chamar> Hi folks, anyone using LXC/LXD?
[04:22] <sdeziel> chamar: yes
[04:23] <chamar> sdeziel, Are you using the apt-get package or snap?!  Having package at both place is kind of confusing
[04:24] <sdeziel> chamar: I'm using the apt-get package because I set this up before the snap came in. That said, upstream recommends using the snap now
[04:25] <chamar> Thanks, I'll go with snap then. Appreciate.
[04:25] <sdeziel> np
[04:27] <Goop> is it possible to enable/disable a graphical desktop environment on-demand?
[04:44] <sarnold> you can certainly apt-get install unity or fvwm2 or whatever you want .. and then apt-get purge it again later
[04:52] <sdeziel> Goop: on a 16.04 desktop, you could set the lightdm service to not autostart (systemctl disable lightdm) and manually start it when you want a graphical session
[04:54] <mason> Goop: you don't mean "systemctl set-default" do you?
[04:55] <mason> ...because if youre going to have a generic system-wide default on Unixlike systems, of course it should really mean "is this my laptop or not". :P
[04:55] <mason> s/default/& setting/
[05:03] <chamar> urgl.. what is that netplan thing?! :/
[05:03] <mason> chamar: You should still be able to install/usr ifupdown.
[05:04] <chamar> yeah.. ho.. that seems to be tied with systemd.. hum
[06:03] <cpaelzer> good morning
[09:50] <rbasak> nacc_: oh. To import the keys into keyrings, yeah, sorry. Any version fo gpg will do for that though.
[11:01] <soshiant> i want to install snmpd
[11:01] <soshiant> apt-get install snmpd
[11:01] <soshiant> errror is : libsnmp30 depends libsensores
[11:02] <soshiant>  i use ubuntu server repository
[11:02] <soshiant> trusty distribution
[11:02] <soshiant> snmpd depends libmysqlclient18
[11:19] <hateball> soshiant: can you !paste the full output?
[11:19] <hateball> soshiant: also run an "sudo apt-get update" first
 libsnmp30 depends libsensores4 but it is not installed
 snmpd depends libmysqlclient18 but it is not installed
[11:41] <hateball> soshiant: can you paste what "apt-cache policy libsnmp30" says?
[11:41] <hateball> !paste | soshiant
[11:41] <soshiant> i use ubuntu server repository
[11:41] <soshiant> i can't past
[11:43] <Ussat> When 18.04LTS goes live, what ver of php will it have ? 7 ?
[11:51] <soshiant> https://paste.ubuntu.com/p/7DrW4n74r5/
[12:53] <frickler> jamespage: you might want to fix your ceph-volume patch to avoid overlap with debian/python-cephfs.install:usr/lib/python2*/dist-packages/ceph_volume_client.py*, see https://git.launchpad.net/~j-harbott/ubuntu/+source/ceph/commit/?id=a874f13042b58073906b9a4465c3498d99301d3c
[13:26] <jamespage> frickler: thanks
[13:29] <soshiantt> .
[14:07] <boxrick> Good afternoon folks, just dabbling with Ubuntu 18 and I would like to drop netplan and use the stock systemd-networkd. Is there any caveats or things to be aware of in this?
[14:07] <boxrick> And a simple way of disabling netplna
[14:07] <boxrick> Netplan
[14:13] <rbasak> boxrick: https://wiki.ubuntu.com/Netplan#I_really_do_need_ifupdown.2C_can_I_still_use_it.3F
[14:13] <rbasak> Oh I'm sorry.
[14:13] <rbasak> That's not exactly what you asked.
[14:14] <rbasak> But I suppose if you used ifupdown in the way described, it's easy enough to configure it to do nothing.
[14:14] <rbasak> I suspect exactly the same applies with netplan - if not given a plan, it will do nothing.
[14:14] <rbasak> cyphermox: ^
[14:15] <boxrick> So lets assume I remove the netplan config then insert the systemd networkd config and restart networkig
[14:15] <boxrick> networking*
[14:15] <rbasak> What does "restart networking" mean? :)
[14:16] <boxrick> Good point, since that doesn't work in Ubuntu 18
[14:16] <rbasak> The term has no meaning.
[14:16] <rbasak> Way back when (before upstart) it might have meant "run the restart action on the networking init.d script)
[14:17] <rbasak> Nowadays with more complex networking setups, and hotplug and so on, it has little meaning.
[14:26] <cyphermox> boxrick: have you filed bugs about whatever doesn't work for you?
[14:26] <cyphermox> and yeah, if you just remove the config, it will do nothing
[14:37] <boxrick> There is no bugs
[14:37] <boxrick> Just rather first testing
[14:37] <boxrick> I need a multi OS networking config, so netplan isn't suitable
[14:37] <boxrick> Configuring systemd-networkd directly is ideal.
[14:38] <boxrick> So as a brief example, if I have eth0 configured by netplan on DHCP. If I remove the config, recreate in systemd-networkd and do a service systemd-networkd restart. Will it spring to life ?
[14:38] <boxrick> Is there any way to confirm netplan is not doing the config to make sure its working.
[14:39] <cyphermox> if there is nothing in /etc/netplan, netplan will do absolutely nothing
[14:40] <cyphermox> but yes, you can use netplan to generate a configuration, then edit it as a starting point for using systemd-network directly -- so once you have config for systemd-networkd in /etc/systemd/network, you can restart systemd-networkd and the network should be managed correctly.
[14:57] <rbasak> nacc_: would you mind avoiding prefixing your branch names with lp references please? It stops me being able to tab complete anything :-/
[14:57] <rbasak> A suffix would be fine.
[15:39] <jamespage> frickler: thanks for that - I've pushed ready for next week when I'll upload
[15:41] <frickler> jamespage: nice, maybe you can also consider https://bugs.launchpad.net/ubuntu/+source/ceph/+bug/1752308
[15:46] <nacc_> rbasak: why does it stop you?
[15:50] <rbasak> nacc: because autocomplete stops after nacc/lp173 and after that I basically have to look up the number every time. From my POV your branches might as well be numbered instead of being named.
[15:51] <nacc> rbasak: right, but for me it's handy to know which branch i need to checkout based upon which bug i'm working on
[15:51] <nacc> rbasak: if it's a suffix, i need to know the <string topic> i chose in order to find it
[15:52] <rbasak> I didn't think you'd find it less convenient. It's your choice I guess :)
[15:52] <rbasak> Personally I can remember the names of a few different things more easily than I can remember the numbers of a few different things :)
[15:52] <nacc> rbasak: i'm the opposite :)
[15:52] <rbasak> (and the numbers are typically all alike)
[15:53] <rbasak> nacc: if you're available, would you like to sync before the standup?
[15:54] <nacc> rbasak: ack
[15:54] <nacc> rbasak: i can join now
[15:54] <rbasak> omw
[15:54] <coreycb> jamespage: down to 3 failures with tempest on queens
[15:55] <coreycb> jamespage: i'm planning to get the final releases uploaded and then work through those last 3 failures
[15:56] <jamespage> coreycb: +1
[15:56] <jamespage> sounds good - do you need a hand?
[15:56] <jamespage> happy for you to take all of the release glory but if you want to spread the load I have some 'waiting for functional tests to complete' time
[15:58] <coreycb> jamespage: either way is fine with me
[15:58] <coreycb> jamespage: should be relatively light-weight at this point
[15:58] <jamespage> coreycb: ok shall I take the neutron* and networking* bits then?
[15:58] <jamespage> and you can work the rest
[16:05] <coreycb> jamespage: sure, sounds good
[16:43] <nacc> rbasak: do you have a few minutes? just wanted to check something with you
[16:43] <rbasak> nacc: sure. Which HO?
[16:43] <nacc> rbasak: standup is fine
[16:44] <rbasak> omw
[16:47] <arunpyasi> Hi everyone, I need to have a secondary mail server which can be used to receive and send emails when primary server is down and the emails should be sync with one another. How do I achieve that ? A good tutorial/manual/tool/link would be appreciated. Thanks !
[16:48] <TJ-> arunpyasi: if using postfix a smarthost config on 1 of them
[16:59] <arunpyasi> TJ-,doing that will sync the emails on both the sides ?
[17:02] <TJ-> arunpyasi: the idea is the SmartHost receives them and passes them on to the primary server when it can
[17:09] <arunpyasi> TJ-, ok, I went through this soln but the thing is, I need each other to be synced and server2 be available for pop/imap/smtp when server1 is down
[17:11] <TJ-> arunpyasi: then you'd need file-system level syncing between them of the mailbox directory/ies so POP/IMAP clients can see the emails
[17:13] <arunpyasi> TJ-, ok so this goes to the filesystem sync ? can it be rsync or GlusterFS ?
[17:13] <Ussat> You need 2 servers that share the same FS, and the config on the servers should reflect that
[17:13] <TJ-> arunpyasi: take your pick of options :) I think I'd want a distributed file-system of some sort to make the sync transparent, but then you hit the problem of, at some times, needing it to switch masters
[17:13] <Ussat> also you need a smart way to determine when one is down and automatically fail to the live server
[17:14] <TJ-> arunpyasi: might be worth looking at how high-available servers are configured for jsut this kind of thing
[17:14] <Ussat> you could use a VIP that then goes to one or the other servers, the VIP could live on something like a F5
[17:14] <Ussat> which would switch from one to the other based on which is live
[17:25] <arunpyasi> Ussat, what is VIP ? AFAIK is Very Important Person
[17:26] <Ussat> Virtual IP
[17:27] <Ussat> So.....in the case of your scenario, the IP would reside on the F5 or similar, and the F5 would forward requests to one or the other server
[17:27] <Ussat> AKA Poor Mans HA
[17:27] <Ussat> Its perfectly viable though and works quite well
[17:28] <Ussat> Hard to explain over text
[17:42] <jamespage> coreycb: ok have neutron* and networking* built - about to publish to bionic once everything is differed
[17:43] <coreycb> jamespage: awesome
[17:59] <jamespage> coreycb: publishing now - I'll need to eod shortly
[18:00] <coreycb> jamespage: ok thanks. i should be able to get the rest today. work is underway on my end.
[18:32] <dpb1> _bladernr_: the bug looks fine.  Thanks.  nacc is working on it now, we'll let you know where there is something to test.
[18:32] <nacc> dpb1: thanks
[18:34] <_bladernr_> dpb1, thanks!
[18:34] <nacc> rbasak: to be sure, you're planning on acking my branch (fixes) before you EOW?
[18:35] <nacc> rbasak: and/or reviewing i guess
[18:35] <dpb1> nacc: good thing you pinged him here. :)
[18:35] <nacc> dpb1: well, it's what we had talked about earlier, but i wanted to be sure -- needed before we can ramp up phasing
[18:38] <dpb1> nacc: understood
[19:11] <rbasak> nacc: yeah I am. I deferred finishing work to do some chores that needed setting off earlier (like laundry). Now back to finish work and OW :)
[19:11] <rbasak> EOW
[19:11] <nacc> rbasak: np, i'll need to reboot in a sec, but i think this snap test should pass now (turns out gpg is senstivie to the order of its options :)
[19:27] <nacc> rbasak: ok, got self-test working with yoru branch
[19:27] <nacc> i'll update mine
[20:17] <nacc> rbasak: fyi, tests pass with my branch, i think i'm ok to land (i could wait for the integration tests, but it's a noop there, afaict)
[22:25] <nacc> powersj: hrm, do you know if the jenkins is perhaps not retriggering right now on branch changes?
[22:25] <nacc> powersj: i just updated ~nacc/lp1734905-script-fixes and i'm not seeing the CI job yet
[22:25] <powersj> nacc: https://jenkins.ubuntu.com/server/job/git-ubuntu-ci-trigger/17270/console that's the latest job
[22:25] <nacc> usually it's pretty immediate
[22:25] <powersj> it runs every 15mins
[22:25] <nacc> oh ok
[22:26] <nacc> maybe i always just got caught it before :)
[22:50] <coreycb> jamespage: everything for queens should be uploaded at this point. will cross-check the report later. still have 3 tempest failures.
[23:15] <driftmonk> hey guys. I’m trying to set e-tag and last-modified headers on my .htaccess but it only seems to work for requests to my subdomains. any ideas? (Wordpress on Ubuntu)
[23:16] <sarnold> driftmonk: iirc the main config delegates to the htaccess files which settings can be modified; perhaps your vhost configs delegate those things, but the "main" configs do not?
[23:18] <driftmonk> sarnold: are you talking about the httpd.conf file?
[23:19] <driftmonk> sorry i haven't touched web servers in a while
[23:20] <sarnold> driftmonk: yeah, and all the files that it includes
[23:22] <driftmonk> mm yes perhaps. I’m just using digital ocean with a Wordpress on Ubuntu platform. where on the machine can I find the configs?
[23:23] <sarnold> /etc/apache* I think
[23:23] <driftmonk> ahh! right! thanks sarnold. will have a look! cheers :)
[23:33] <pmatulis> how can i discover whether a certain package is installed by default on an ISO? for instance, how to discover whether 'lxd' is installed on Artful Server or Artful Desktop?
[23:33] <nacc> pmatulis: seeded-in-ubuntu command
[23:34] <nacc> pmatulis: and 'preinstalled' in the output for server or destkop as appropriate
[23:36] <patdk-lap> never installed for me
[23:36] <patdk-lap> but that is cause I only install jeos
[23:39] <pmatulis> nacc, sweet! but this covers only the running release right?
[23:41] <pmatulis> also, for 'snapd', i see 'ubuntu-server: daily, daily-live, daily-preinstalled'
[23:41] <pmatulis> what about the official ISO? and there are live ISO's for server?
[23:42] <nacc> pmatulis: i'm actually not sure, i think it reflects the current seeds (e.g., for bionic), but i really don't know
[23:42] <nacc> pmatulis: there is a new installer iso now (subiquity)
[23:42] <nacc> pmatulis: i wonder if that is waht the daily-live is
[23:43] <nacc> powersj: --^ maybe you can answer better?
[23:43] <mwhudson> er i don't think so
[23:43] <mwhudson> the live-server is just the server seed + sprinkles
[23:43] <nacc> oh ok
[23:44] <mwhudson> i think, anyway :)
[23:46] <mwhudson> or is it the cloud-image seed?
[23:54] <mwhudson> minimal, standard, server and cloud-image _tasks_ it seems
[23:54] <nacc> mwhudson: ah