[00:05] <rizonz> sarnold: checkout debmirror, much nicer
[00:05] <rizonz> for ubuntu likes
[06:13] <lordievader> Good morning
[07:05] <cpaelzer> rbasak: if you ever wonder, those squid tests are run-once by design
[07:05] <cpaelzer> rbasak: I don't want you to waste the same 30 minutes I just wasted
[07:06] <cpaelzer> rbasak: they end with adding apparmor profiles that won't work for certain aa-tests, a rerun fails until you unload the profile
[07:11] <sarnold> o_O
[07:21] <col88> Hi! im trying to add an entry to /etc/hosts using preseed - any tips?
[07:24] <cpaelzer> col88: you can at least force your own hostname into /etc/hosts with netcfg/hostname https://help.ubuntu.com/lts/installation-guide/example-preseed.txt
[07:24] <cpaelzer> col88: but I guess you want more hosts into /etc/hosts ß
[07:24] <cpaelzer> ?
[07:25] <col88> essentially i need to tell my new hosts that my puppetmaster is identified as IP address xxxx
[07:25] <col88> so resolve the name to a static IP on all machines
[07:26] <cpaelzer> col88: sorry, no idea for normal preseeds, but if in your env cloud-init is an option http://cloudinit.readthedocs.io/en/latest/topics/examples.html#run-commands-on-first-boot
[07:27] <col88> ah ha thats a good idea
[07:45] <rbasak> cpaelzer: noted, thanks!
[11:01] <cpaelzer> rbasak: it turned out most fixes that I was writing were already in the upstream qa tests
[11:01] <cpaelzer> rbasak: so I pulled the latest version of those
[11:01] <cpaelzer> rbasak: and on top only tweaked to run against localhost
[11:01] <cpaelzer> rbasak: and a bit of setup to have something reasonable available at localhost
[11:02] <cpaelzer> rbasak: my tests on LP are starting soon (building atm), once successful we could push to z.p
[11:02] <cpaelzer> rbasak: although we are in freeze and this is not 100% bug-fix of the package itself
[11:02] <cpaelzer> rbasak: but I think proper testing would still be welcome
[11:03] <cpaelzer> rbasak: before that the tests have to actually pass on LP
[11:03] <cpaelzer> rbasak: to confirm - those tests are pure Ubuntu Delta atm, no need to send to Debian or did you intend to do so once they are working fine?
[11:06] <rbasak> cpaelzer: I intended to send to Debian once they were in reasonable shape, as I didn't think the previous state of them would be acceptable to Debian.
[11:06] <rbasak> cpaelzer: as for the freeze, I think dep8 fixes are absolutely fine.
[11:06] <rbasak> Or any kind of dep8 improvement, not just fixes.
[11:06] <cpaelzer> ok
[11:48] <blackflow> Hello. Need help understanding the bug resolution life cycle. What's the next step for Bug #1673357 ? It's been recommended to the server team, but it'll have to be patched by the maintainers first?
[11:51] <rbasak> blackflow: it's definitely fixed in Zesty in 2.0.33, right?
[11:52] <rbasak> blackflow: in that case see https://wiki.ubuntu.com/StableReleaseUpdates and in particular https://wiki.ubuntu.com/StableReleaseUpdates#Procedure
[11:52] <rbasak> blackflow: anybody can do the preparation steps - the only thing you won't be able to is the actual upload, for which you'll need a sponsor.
[11:52] <rbasak> blackflow: the server team will get round to it eventually I hope, but if you'd like to speed up the process, please do as much as you are able.
[11:54] <blackflow> rbasak: I've been maintainer of some packages on FreeBSD (and bugzilla triager) so I'm in general familiar with the whole contribution process. However, I'd need help getting started with Ubuntu contribs. Is there a process/team/mentorship I can rely upon for that?
[11:54] <rbasak> blackflow: I wish we were better on the documentation front. It's mostly out of date :-/
[11:54] <rbasak> blackflow: feel free to ask here, or in #ubuntu-devel
[11:55] <rbasak> There's https://wiki.ubuntu.com/SponsorshipProcess and some other related pages on the wiki but I'm afraid some of it isn't current.
[11:55] <rbasak> In short, post a debdiff to the bug, and subscribe ~ubuntu-sponsors, and it'll be in the sponsorship review queue.
[11:56] <rbasak> blackflow: if there's a specific thing you feel that we're missing in the process to help get new contributors on board, please let us know.
[11:56] <rbasak> We're working on a git-based workflow for contributions, but that's still a work in progress and not quite ready for general use.
[11:56] <rbasak> When it's ready, I'm hoping that we can replace most of the documentation with just "clone from here, push to here, submit a merge proposal here" flow.
[11:57] <rbasak> blackflow: in the meantime, I'd be happy to answer specific questions.
[12:01] <blackflow> rbasak: thanks. :) I don't have any for the moment, and in fact you've given me a pretty sound advice with debdiff.
[13:37] <zul> jamespage: question for you do we really need this patch? https://git.launchpad.net/~ubuntu-server-dev/ubuntu/+source/keystone/tree/debian/patches/add-version-info.patch
[14:11] <jamespage> zul: :-)
[14:12] <zul> umm..yeah
[14:17] <cpaelzer> rbasak: grr - of course now other tests fail https://objectstorage.prodstack4-5.canonical.com/v1/AUTH_77e2ada1e7a84929a74ba3b87153c0ac/autopkgtest-zesty-ci-train-ppa-service-2667/zesty/amd64/s/squid3/20170329_132948_8ac14@/log.gz
[14:18] <rbasak> :-/
[14:19] <cpaelzer> maybe I updated too much believing in the good of "new" too much
[14:19] <cpaelzer> but hey, http and https worked in LP infra
[14:19] <rbasak> cpaelzer: could it be because a proxy is set by default?
[14:19] <rbasak> Not sure about FTP though.
[14:19] <rbasak> But that sort of thing has been a cause of failures in the past.
[14:20] <rbasak> Causing requests intended for localhost to bounce through some external host that then resolves "localhost" differently, etc.
[14:22] <cpaelzer> rbasak: ah no I think I see the issue
[14:22]  * cpaelzer is touching all kind of wood
[14:25] <eanyx>  Hi, I need to make a cluster with an apps that was not designed to be in cluster. I would like to use keepalive + haproxy for failover. But how to watch for apps failover?
[14:27] <eanyx> I've a problem : I need to have rabbitmq writing from the 2 nodes on the same shared queue
[14:31] <cpaelzer> eanyx: doesn't rabbitmq have heartbeat and clustering as feature ? https://www.rabbitmq.com/reliability.html
[14:32] <cpaelzer> rbasak: that was good to fail actually
[14:33] <cpaelzer> rbasak: it had still some ubuntu.com references which might work for us but not on dbci
[14:33] <cpaelzer> rbasak: and the issue with those tests helped me to uncover
[14:33] <cpaelzer> rbasak: a new revision is building, but it builds oh-so-slowly
[14:33] <rbasak> Ah, great!
[14:33] <cpaelzer> rbasak: I can only trigger tests later on after some family life
[14:33] <rbasak> Sure :)
[14:33] <eanyx> yes, but how to integrate with my apps? I need some failover software to switch my apps from one node to other and then use anoter clustering techno (rabbitmq to switch over)?
[14:34] <cpaelzer> sorry eanyx, I just don't know - crossing fingers for you that somebody else does
[14:35] <eanyx> cpaelzer: are you aware of reentrant software?
[14:38] <cpaelzer> sure, but from reentrant functions via acid on a DB to no-state containers that can be something different for each of 10 people you ask
[17:08] <rizonz> sarnold: ping
[17:23] <sarnold> hey rizonz, how's your mirroring project coming?
[18:01] <cpaelzer> rbasak: https://bileto.ubuntu.com/excuses/2667/zesty.html all good now on squid
[18:02] <cpaelzer> rbasak: I'll do the publish and Debian submit tomorrow
[20:13] <rbasak> cpaelzer: o/ thank you1
[20:13] <rbasak> !
[23:28] <rizonz> is it possible to use metalink files for yuour local mirror ?
[23:30] <drab_> rizonz: did you figure out why your mirror kept getting bigger and bigger?
[23:31] <rizonz> drab_: he thanks for asking, dunno, I really cannot say... debmirror works perfectly
[23:37] <drab_> rizonz: so you were using apt-mirror before?
[23:37] <drab_> I actually need to do something about it because it's really slow and not allowing me to resync things selectively, at least not that I can find out
[23:38] <drab_> so if I have 5 things to mirror I can't just kick one off (unless I temporarily mess with the config file)
[23:40] <rizonz> drab: yap
[23:40] <rizonz> drab: it's damn slow
[23:41] <rizonz> debmirror did the erm, 150G in 1,5H :)
[23:41] <drab> wow, ok
[23:42] <drab> rizonz: the only thing is, debmirror seemed a lot more complicated to setup/maintain compared to apt-mirror, but maybe I just looked at the wrong howto
[23:42] <rizonz> drab: sec, have a doc for you... it's damn easy
[23:44] <rizonz> drab: https://help.ubuntu.com/community/Debmirror
[23:47] <drab> rizonz: yeah, hacky :) . compared to adding the lines to mirror to a config file, that is
[23:47] <drab> rizonz: apt-mirror is a lot more consistent with the whole apt setup, configs are even in /etc/apt/
[23:48] <drab> so it's overall a lot more "sensible", but yeah, can't seem to be ablet o get it to work well so will need to ditch it
[23:48] <rizonz> drab: apt-mirror sucks
[23:48] <rizonz> comparing to debmirror
[23:49] <rizonz> drab: I don't think this is hacky
[23:50] <drab> that's fine, we clearly have diff definitions of hacky, not a problem :)
[23:50] <rizonz> drab: if you want a full mirror, just rsync it
[23:50] <rizonz> I do that with fedora as they repo structure it MUCH better
[23:51] <rizonz> *their
[23:51] <drab> that has other problems, no logs, which is also why I don't like backups with plain rsync + hardlinks even tho conceptually it's a good way to go, minus deduplication on large files/vm images
[23:59] <rizonz> drab: no just a good exclide file ;)
[23:59] <rizonz> *exclude