[01:15] <wallyworld> and here's that forward port PR we discussed https://github.com/juju/juju/pull/11320
[01:28] <thumper> moar coffee needed
[01:52] <babbageclunk> I'm having trouble bootstrapping in lxd - get an error that it can't get mongodb-server-core. Anyone else seeing that?
[01:53] <thumper> babbageclunk: which series?
[01:53] <babbageclunk> bionic
[01:53]  * thumper tries
[01:55] <thumper> command is running, I'll let you know when it's done
[01:56] <babbageclunk> I can install it fine on a fresh lxc instance
[01:56] <tlm[m]> got it running also, no issues so far
[01:56] <babbageclunk> I'm trying again
[01:58] <babbageclunk> :(
[02:00] <thumper> babbageclunk: new lxd controller started fine
[02:00] <thumper> as did enable-ha on an existing controller
[02:04] <tlm[m]> same problem babbageclunk , fails to install mongo
[02:06] <tlm[m]> seems problem is with package libyaml-cpp0.5v5 not existing
[02:22] <babbageclunk> tlm[m]: It looks like apt update is failing for me (with a name resolution error) and then everything else is a consequence of that.
[02:24] <tlm[m]> no seeing any dns issues in my logs. Looks to be package missing for me
[02:24] <tlm[m]> not*
[02:32] <babbageclunk> tlm[m]: do you have squid-deb-proxy set up? maybe mine's gone rogue
[02:32] <tlm[m]> Nup
[02:40] <babbageclunk> yay worked it out
[02:40] <thumper> oh?
[02:42] <babbageclunk> I had an apt-http-proxy in my .local/share/juju/clouds.yaml entry for lxd, and I guess that proxy went away recently?
[02:42] <babbageclunk> (wasn't to a local address)
[02:43] <babbageclunk> I think it was the VPN squid, added at some point when there was a problem
[02:43] <babbageclunk> and then not removed when the problem was fixed later.
[02:43] <thumper> :)
[04:08] <babbageclunk> can I get a review for a backport? https://github.com/juju/juju/pull/11322
[04:18] <wallyworld> babbageclunk: looking straight after meeting
[04:18] <babbageclunk> thanks
[04:26] <wallyworld> babbageclunk: lgtm ty
[04:26] <babbageclunk> wallyworld: thanks!
[10:52] <flxfoo> Hi all,
[10:54] <flxfoo> I would like to upgrade my percona cluster. So after pausing ha and non-leader member, I would upgrade and reboot. Then resume each of those non leader. Should I switch leader to upgrade the last one ? How would I do that?
[12:30] <stickupkid> manadart, you ever seen this
[12:30] <stickupkid> https://paste.ubuntu.com/p/TXd3pyfTvh/
[12:33] <manadart> 2:15:58 INFO  juju.environs.manual.sshprovisioner sshprovisioner.go:43 initialising "10.132.183.234", user ""
[12:33] <manadart> I think the endpoint needs to be user@addr.
[12:33] <manadart> Maybe.
[12:33] <stickupkid> documentation fail
[12:35] <stickupkid> manadart, fixed that, but still fails https://paste.ubuntu.com/p/rDpCVmQvcd/
[12:35] <stickupkid> manadart, i'll have a look later
[12:36] <manadart> stickupkid: Try explicitly adding the key that you know will work to the SSH agent...
[13:54] <stickupkid> manadart, https://github.com/juju/juju/pull/11317/files
[14:32] <flxfoo> hi again
[14:33] <flxfoo> `juju add-machine` does not find the image, how can change the image ID?
[15:12] <stickupkid> manadart, I know this a draft and only targetting bootstrap, can you take a quick look to see if you agree before I move on to deploying path way
[15:12] <stickupkid> manadart, might help if I add the PR https://github.com/juju/juju/pull/11324
[15:22] <manadart> stickupkid: Commented. Looks OK.
[15:24] <stickupkid> manadart, I'm struggling to workout the right thing for deploy as it serves two masters
[16:19] <stickupkid> manadart, got a sec
[16:19] <stickupkid> ?
[16:19] <manadart> stickupkid: Daily?
[16:20] <stickupkid> yeap
[16:53] <hml> anyone know the difference in the branches for the go sys pkg:  release-branch.go1.14-std vs release-branch.go1.14-cmd
[20:03] <hpidcock> hml: https://github.com/golang/go/commit/866920a07351ad5663cc712c31a0f7b5631bc85c
[20:03] <hpidcock> std is what we want
[20:03] <hpidcock> cmd is for the go tools
[20:04] <hml> hpidcock:  given that we have to still compile with 1.11 and 1.12 for the near future.  moving pieces up to 1.14 seems problematic on all architures
[20:05] <hml> hpidcock:  ty for the clairification
[20:06] <hpidcock> hml: well then we should use the 1.12 pkgs
[20:06] <hpidcock> but 1.14 pkgs for the most part should work on 1.10
[20:10] <hpidcock> hml: actually 1.12 is no longer supported, so I think it's received it's final update
[20:10] <hpidcock> ideally we use 1.13 pkgs if possible.
[20:11] <hml> hpidcock:  we really need to get the darn releases off of the early versions, that’s the hold up in my mind
[20:11] <hml> we can build all we want with new stuff… but if the release or the snap is foobar… it doesn’t matter
[20:11] <hml> i know it’s in the works.
[20:12] <hml> s/release/ppa
[20:12] <hml> we have a bunch of other golang.org/x/* dependendies pinned to old stuff.  not just sys
[20:13] <hpidcock> problem is we need to bump x/crypto
[20:14] <hml> hpidcock:  i’ve had trouble ensuring that just the sys pkg update will resolve the issue with s390x. i’m getting i/o timeouts on make rebuild-dependencies
[20:14] <hpidcock> :/
[20:14] <hml> hpidcock:  is a no win right now. with cypto and the bug fix
[20:41] <hml> hpidcock:  review pls https://github.com/juju/juju/pull/11256
[20:44] <hpidcock> hml: done
[20:44] <hml> hpidcock: ty
[23:32] <timClicks> babbageclunk: have a minute to chat vsphere?
[23:32] <babbageclunk> timClicks: sure, in standup?
[23:32] <timClicks> um yip 1 min