[09:16] achilleasa, bringing in the changes for the charm repo around service-less bundles https://github.com/juju/charmrepo/pull/165 [09:53] achilleasa, https://github.com/juju/bundlechanges/pull/75 [11:02] achilleasa, finally brings in the changes for service-less bundle yaml https://github.com/juju/juju/pull/12482 [11:02] I'm expecting test failures somewhere, I can feel it in my bones [11:07] stickupkid_: do we know whether any of the existing CS/CH charms have service blocks? [11:22] achilleasa, they shouldn't do... [13:43] achilleasa, another https://github.com/juju/python-libjuju/pull/459 [13:57] stickupkid_: trade you for https://github.com/juju/juju/pull/12483 [13:59] why python 2? [14:00] I got the usual regex-related issue when running under python3 so I assumed these are running with 2 :D [14:01] I could change them to be python3 friendly though [14:01] not sure how many tests will be affected [14:01] python3 only [14:01] yeah, we'll have to do the work at somepoint [14:01] sorry :( [14:02] ok, I will make it work with python3 and see if CI explodes after I land this :D [14:02] haha [15:11] Hey good people, when I try to take a juju controller backup, it fails with error : Failed: error writing data for collection `logs.logs.bac50024-0ebc-4409-8261-2cf17197e703` to disk: error reading collection: Executor error during find command: CappedPositionLost: CollectionScan died due to position in capped collection being deleted. Last seen [15:11] record id: RecordId(24562884); [15:12] As per google, this normally happens in mongodb when new data has overwritten the next document so the scanner couldn't continue [15:12] Is there any way to overcome this? [15:13] This does not happen all the time. [15:13] There are cases when the juju controller backup is success