[03:07] <mwhudson> is it my imagination or is the ppa publisher running much faster than it used to?
[03:07] <cjwatson> mwhudson: Probably not your imagination; William did a big data optimisation pass recently.
[03:10] <mwhudson> well hooray for that
[03:10] <mwhudson> cjwatson: also go to sleep!
[03:15] <cjwatson> That would be lovely.
[03:15] <cjwatson> Sadly physiology.
[03:19] <wgrant> mwhudson: Publish latency is usually below 10 minutes now, whereas before October it was frequently >40 minutes. Various bits of optimisation gave fairly impressive results.
[06:27] <mwhudson> wgrant: is it better for the archive too, or just ppas?
[06:27] <mwhudson> but still >10 min publisher latency for ppas is awesome
[06:27] <mwhudson> er *<10 min, obviously
[06:33] <wgrant> mwhudson: The Ubuntu primary archive is still slow, because a-f on an Ubuntu-sized suite takes >20min
[06:33] <mwhudson> wgrant: ack
[06:33] <mwhudson> wgrant: assuming it doesn't OOM the machine :)
[06:34] <wgrant> mwhudson: That wasn't a-f!
[06:34] <wgrant> a-f was a victim.
[06:34] <wgrant> Possibly the only time ever that a-f has been innocent of anything, but, well.
[06:34] <mwhudson> well yes, i guess so :)
[06:35] <mwhudson> wgrant: is there much that could be done to speed a-f up? i presume it's a pile of perl
[06:35] <mwhudson> huh no, apparently not
[06:46] <acheronuk> wgrant: I noticed that better publishing time, and very much appreciate the work that went in to to that by people :)
[07:56] <acheronuk> wgrant: Can the size our CI ppa be bumped a bit please? https://launchpad.net/~kubuntu-ci/+archive/ubuntu/unstable/+packages
[07:57] <acheronuk> We now do 3 releases on that, and additional Qt etc required for that has just tipped it over the limit by the looks of things
[08:01] <wgrant> acheronuk: Doubled.
[08:01] <acheronuk> wgrant: Thank you :)
[18:17] <marcoceppi> wgrant: hey, we also need a PPA size increase for ppa:juju/stable, we've run out of space :(
[18:48] <cjwatson> marcoceppi: Doubled, but do you actually need to keep e.g. packages for quantal?
[18:48] <marcoceppi> cjwatson: probably not, I was planning to go through and clean up everything not trusty, xenial, yakkety, z
[18:48] <marcoceppi> cjwatson: thanks though
[18:48] <cjwatson> Maybe precise too ...
[18:48] <cjwatson> (I mean, maybe keep precise.  It's not EOL yet)
[18:49] <marcoceppi> ah, yes