[00:15] <xnox> vorlon:  I am failing to test my latest edition of resolved integration with ifupdown & isc-dhcp by upgrading xenial to groovy. Because of invalid shell syntax in ifenslave. I don't think diff conflict markers are valid in dash sh.
[00:15] <xnox> https://git.launchpad.net/ubuntu/+source/ifenslave/tree/debian/ifenslave.if-pre-up?h=applied/ubuntu/devel#n129
[00:15] <xnox> vorlon:  do you want to fix it up? or should i? can we just sync ifenslave at this point?
[00:15] <xnox> (forcesync)
[00:15] <xnox> vorlon:  it is orphaned in debian, so we can just upload our delta there.
[00:42] <vorlon> "diff conflict markers" well then
[00:43] <vorlon> xnox: I don't know than any of these changes are correct in Debian where ifupdown still exists
[00:46] <vorlon> xnox: I'd rather we just get back in sync with Debian as-is given that this is a universe package that depends on ifupdown
=== tmhoang4 is now known as tmhoang
[05:23] <cpaelzer_> sergiodj: there must have been two libpcap uploads even we couldn't find the first one yesterday
[05:24] <cpaelzer_> bdmurray: was so kind to SRU-accept one and cancel the other
[09:34] <niub> o/ is there a place where I can see all the bug related to a specific kernel version?
[09:34] <niub> I think there is a bug in 4.15.0-109.110 related to btrfs
[11:03] <rbasak> jamespage: o/ I'm looking at openvswitch in the SRU queue. Looks like your Xenial upload got trumped by a security update, sorry.
[11:05] <rbasak> Looking at Bionic now
[11:06] <jamespage> rbasak: no worries - I'll rebase
[11:06] <jamespage> thanks
[11:07] <jamespage> rbasak: bdmurray pinged me about the duplicate uploads as well
[11:07] <jamespage> there more recent one is not based on a git snapshot so the delta is smaller
[11:07] <rbasak> I spotted your comment in the bug and rejected the older one as requested
[11:07] <rbasak> Thanks!
[11:08] <jamespage> brill thankyou
[11:14] <rbasak> jamespage: where did you get the new orig tarball from? It seems to mismatch https://www.openvswitch.org/releases/openvswitch-2.9.6.tar.gz - different compression, but also the raw tar
[11:24] <jamespage> rbasak: d/copyright triggers a repack
[11:24] <jamespage> gz->xz and exclusion of some files
[11:29] <rbasak> Ah - thanks!
[11:30] <rbasak> I can figure it out from there then. Seems odd and maybe wrong though - no dfsg label, and only debian/ excluded? Isn't that automatic with 3.0 (quilt) anyway?
[11:30] <rbasak> To be clear, I'm not bothered about that for the SRU - just thought I'd mention it
[11:33] <xnox> vorlon:  looking at the code, what we have in 2.9ubuntu1 de-races bringing multiple bonds up. I do not see anything remotely close to fixing that in debian. I've uploaded a revert back to 2.9ubuntu1 as 2.10ubuntu2
[11:33] <xnox> vorlon:  cause i'd rather keep bonds working the same way they did in focal and down.
[11:34] <xnox> vorlon:  hm, but i should drop the ifenslave binary i guess.
[11:51] <rbasak> doko: on bug 1888386, shouldn't these go via the security pocket?
[11:51] <ubottu> bug 1888386 in openjdk-14 (Ubuntu) "SRU: backport the openjdk-13 and -14 security releases to focal" [Undecided,New] https://launchpad.net/bugs/1888386
[12:14] <doko> rbasak: when I talked with vorlon and sil2100, they preferred the updates pocket. but I don't mind
[13:55] <sergiodj> cpaelzer_: that's great, thanks bdmurray :)
=== cpaelzer_ is now known as cpaelzer
[14:12] <xnox> mwhudson:  it would be nice if you use -x option to git cherry-pick when cherrypicking things. I.e. it's hard to see in livecd-rootfs bionic, which commits have been cherrypicked and which ones are still missing.
[16:09] <GunnarHj> Hey sil2100! The bionic langpack testing deadline is reached:
[16:09] <GunnarHj> https://wiki.ubuntu.com/Translations/LanguagePackUpdatesQA
[16:25] <sil2100> GunnarHj: oh well, it's better than what I expected anyway
[16:25] <sil2100> Thanks!
[16:25] <sil2100> I'll publish those a bit later
[16:26] <GunnarHj> sil2100: Yeah, at least some of the most important languages were tested.
=== ijohnson is now known as ijohnson|lunch
[19:11] <ahasenack> sil2100: hi, if you are around, looks like bileto doesn't like it when a binary package is no longer built: https://bileto.ubuntu.com/excuses/4170/groovy.html
[19:11] <ahasenack> in this case, it's a transitional package I can finally drop
[19:12] <ahasenack> "missing build on all: libpmemobj-doc (from 1.8-1ubuntu1) "
[19:12] <ahasenack> can that check be ignored?
[19:39] <mwhudson> xnox: til
[20:31] <Laney> ahasenack: I think https://git.launchpad.net/bileto/tree/britney/britney.conf.in needs to get https://git.launchpad.net/~ubuntu-release/britney/+git/britney2-ubuntu/tree/britney.conf#n32 options ALL_BUILDARCH and HAS_ARCH_ALL_BUILDDS
[20:31] <Laney> if you wanted to submit a merge proposal that would probably be welcomed :>
[20:32] <ahasenack> I have no idea how britney works
[20:32] <ahasenack> but let's see
[20:33] <Laney> without those options britney thinks there's an arch called 'all' which can fail on its own
[20:34] <Laney> so I think it's going "omg where are the binaries for 'all' which I knew about before, I don't have any"
[20:39] <ahasenack> Laney: this? https://pastebin.ubuntu.com/p/DPB45jnRf6/
[20:40] <Laney> ahasenack: looks right to me
[20:42] <ahasenack> hmpf, I miss a "fork" buttin in lp
[20:42] <ahasenack> butotn*
[20:43] <ahasenack> now to figure out the push url
[20:44] <ahasenack> ok, got it
[20:46] <ahasenack> done
[20:47] <Laney> I'd merge if I could!
[20:50] <ahasenack> :)
[20:50] <ahasenack> thanks! :)
[20:50] <ahasenack> I also don't know if there is a separate deployment step
[20:51] <Laney> me neither
[21:40] <cjwatson> fork button> it's coming!
[22:20] <mwhudson> is the huge build queue fallout from the outage or did the archive rebuild get started?
[22:20] <mwhudson> actually to be that huge it must be the rebuild
[22:25] <cjwatson> mwhudson: Test rebuild, yes
[22:25] <mwhudson> cjwatson: ta
[22:25] <cjwatson> The outage should be cleared, and so far fixed in a couple of different ways ...
[22:28] <cjwatson> (Cowboyed code fix for my stupid max/min mistake; code fix under review to move the slow bit of database interaction to a better place; configuration change to have fewer concurrent download connections; code fix under review to make the librarian client use epoll rather than select so that it doesn't matter if the download process pool has lots of FDs open; and some OpenStack thwacking for ...
[22:28] <cjwatson> ... ppc64el builder resets)
[22:28] <cjwatson> Test rebuilds are good at shaking out this sort of problem
[22:30] <cjwatson> Some of this has probably been latent for a while and causing low-level problems, just not quite enough to track down and fix
[22:49] <mwhudson> cjwatson: nothing like fixing bugs several times over