[13:16] <smoser> o/
=== mgerdts_ is now known as mgerdts
=== r-daneel_ is now known as r-daneel
[15:20] <blackboxsw> minor nits on https://code.launchpad.net/~james-hogarth/cloud-init/+git/cloud-init/+merge/333657
[15:21] <blackboxsw> oops I meant on https://code.launchpad.net/~smoser/cloud-init/+git/cloud-init/+merge/342007
[16:08] <blackboxsw> smoser: rharper the artifacts that jenkins is keeping around for a proposed run are 3.4 GB. The heavy hitters seem to be that we are keeping around qcow images in our artifacts. Think we can prune these ? https://jenkins.ubuntu.com/server/view/cloud-init/job/cloud-init-integration-proposed-a/lastSuccessfulBuild/artifact/
[16:09] <rharper> blackboxsw: I'd think so
[16:09] <rharper> in curtin, we purge images if we succeeed on only keep on failure
[16:09] <rharper> also, you really should tar any qcow2 file you capture as the copy on jenkins isn't sparse aware and will fill up master
[16:10] <blackboxsw> yeah I'd like to be able to attach the artifacts zip to cloud-init's SRU, but attaching 3.X gig file seems a bit irresponsible
[16:10] <smoser> blackboxsw: i think we want to not collect qcow2 and .img files
[16:11] <blackboxsw> +1 I'll put up a PR for jenkins-jobs
[16:12] <blackboxsw> to prune that action
[16:12] <blackboxsw> s/action/bloat
[16:12] <smoser> so.. out of thall those cloud-test-* files..
[16:12] <smoser> i'd like to keep the console log
[16:12] <smoser> they others i'd like to ditch
[16:13] <smoser> actually the console logs are saved in the test output
[16:13] <smoser> so dont need themn from the top level
[16:13] <blackboxsw> yeah was just discovering that too.
[16:14] <smoser> let me see if we can put those files elsewhere so that they're not collected.
[16:14] <rharper> https://bugs.launchpad.net/netplan/+bug/1664844
[16:14] <ubot5`> Ubuntu bug 1664844 in netplan "No distinction between link-up and link-down interfaces" [High,In progress]
[16:15] <rharper> we now have a 'optional' boolean to determine if wait-online cares about an interface
[16:15] <rharper> I think any control-manual interfaces should get optional: False , but need to think some more
[16:16] <smoser> well, reversed, but yes
[16:16] <smoser> control-manual is optional: True
[16:18] <rharper> ah, right
=== nacc_ is now known as nacc
[16:27] <blackboxsw> smoser: rharper https://github.com/canonical-server/jenkins-jobs/pull/33 ?
[16:28] <blackboxsw> this should avoid collecting that artifact for us in jenkins (we'd still be able to look through it if we connect to jenkins cli
[16:34] <rharper> blackboxsw: smoser: ok, updated the ntp branch with something I think covers what we need, the unittest breakage was minimal;  smoser  in particular in the ntp branch, I wanted to get your eyes on the decode_text() changes, it appears that we need to decode the template contents before passing to jinja for rendering on python2.7; so I've changes to handle that but definitely want your comments there
[16:55] <blackboxsw> good deal rharper
[16:55] <blackboxsw> checking
[17:04] <smoser> blackboxsw: so 'data_dir' as we use it is 98% "results dir"