[08:33] moo-ning :-) [09:07] mo'in [09:12] Chipaca: would a 1:1 around 12:30 be fine with you . If so I can add it to your calendar ? [09:21] bthomas: go for it [09:21] thanks [09:36] my day just got 78% better \o/ (a 1h meeting was canceled) [09:53] :-) [10:36] ¡Muy buenos días a todos! [10:36] नमस्ते facubatista [10:37] hola bthomas [10:47] PR operator#435 closed: Only set sys.breakpointhook in Framework.__init__ if env is set [10:48] facubatista: ｇｏｏｏｏｏｏｏｏｏｏｏｏｏｏｏｏｏｏｏｏｏｏｏｏｏｏｏｏｏｏｏｏｏｏｏｏｏｏｏｏｏｏｏｏｄｍｏｒｎｉｎｇ！ [10:49] facubatista: should i merge my suggestions into #432 and land it? [10:49] PR #432: More docstrings (and some improved a little) [10:49] Chipaca, please [10:49] Chipaca, I'll work on other docstrings, unless you're on them [10:50] facubatista: i'm on the fix for #421 [10:50] Issue #421: breakpoint handlers unconditionally installed on framework initialization [10:52] ack [10:53] facubatista, let me know when you landed 432, thanks [11:58] PR operator#432 closed: More docstrings (and some improved a little) [11:58] facubatista: ^ [11:58] cheated a little because windows was taking ages [12:11] Chipaca, in what sense you cheated? [12:11] facubatista: ☐ Use your administrator privileges to merge this pull request. [12:12] Chipaca, ah, it's fine [12:12] yup [12:31] * bthomas pings Chipaca [12:31] whoops [12:31] 1 sec [12:58] * bthomas thanks Chipaca [12:58] de nada 🙂 [13:05] Chipaca, bueno, 424 should be green, and I +1ed 436 [13:05] that's all you should need for today from my side in the OF [13:05] I'll start the charmcraft release process in 5' or so [13:11] facubatista: there'll be one more for OF, i'll ping you [13:11] bthomas: please review #436 if you can [13:11] PR #436: Give ObjectEvents a useful __repr__ [13:12] will do [13:19] Chipaca, sure [14:10] bthomas: I've created a few PRs against the prometheus operator, fwiw - https://github.com/canonical/prometheus-operator/pulls [14:10] mthaddon: will review thank you [14:10] thx - no hurry obviously [14:13] * Chipaca pokes travis [14:16] Good morning o/ [14:17] hola justinclark [14:24] bthomas - just reviewed the Prometheus PRs and they should be good to merge. [14:24] :-) I was just going to request that from you. [14:24] will merge now [14:24] thanks [14:25] I liked the use of assertLogs I think I am going to stick with them in the future [14:33] Agreed ^ I'm going to refactor some of the Grafana tests to use it [14:35] good morning justinclark [14:36] monin JoseMasson [14:36] :-) [14:36] :-) [14:37] I enjoy the friendliness in the morning over here on IRC [14:37] * bthomas wonders why justinclark does not enjoy friendliness at 2AM EST [14:38] * bthomas takes a coffee break [14:40] * justinclark wishes he enjoyed waking up at 2AM but, alas, does not for some strange reason [14:46] * bthomas back [15:16] * facubatista lunches while snap remote building works [15:45] Issue operator#438 closed: defered events never re-emitted when using juju storage backend [15:45] PR operator#439 closed: Make JujuStorage.notices respect the api :-) [15:46] facubatista: does the charmcraft snap include ops? and if so, wouldn't it make sense to release it after we release ops? [16:01] * Chipaca wonders if travis is somehow immunte to being stared at [16:01] immune* [16:11] Chipaca, oh, maybe [16:15] Chipaca, so, I aborted the remote build, I will re-trigger it when you're done [16:16] facubatista: travis is taking very long today [16:16] facubatista: i've got one more pr to push (the rework of the breakpoint init) but i'll push it once the current one is in to minimise builds [16:16] it's straightforward, but i'm not sure it'll make it in before my EOD [16:17] Chipaca, we can release tomorrow [16:18] that's what _she_ said! [16:18] (my dog) [17:22] Before I spend too many brain cycles rearranging charm logic, is there some way in OF to get the current application ingress-address without having a peer relation? [17:26] I believe it's exposed by Juju (it's also an environmental variable in the application pod) but I haven't been able to figure out how to get it through the charm. [17:32] justinclark: would that be part of the juju-info implicit peer relation? [17:36] justinclark: hopefully jam Chipaca or facubatista would be able to help you with your Q. I would be interested to know too. What I was trying to say was this : updating of minimum master nodes is only done in response to a peer relation joined event. Hence by the time you need to update it you may already have a relation to get the ingress address from. Besides operator framework does allow you to check if you have a relation or not and [17:36] also the number of units in the relation. This leads me to believe it may be possible to find an implementation that works even if at present there is no way to get the ingress address without a realtion. Having said so I may in all possiblity be wrong and would gladly stand corrected, if so. [17:42] Chipaca, I will look into juju-info. Is this still how it is used in OF? https://github.com/juju-solutions/interface-juju-info [17:42] justinclark: maybe https://github.com/canonical/operator/issues/395 is a better place to start? [17:43] Thank you Chipaca [18:03] bthomas, I think you're right - we can do this without needing ingress-address. [18:04] I still do want to make sure we have a good solution if there's ever a case where we need to set dynamic settings before a peer joins. But I think this can wait. [18:19] * bthomas is craving a five guys veggie burger and fries [18:19] * bthomas goes === dylan is now known as Guest95360 [19:19] I am having an issue with the test framework [19:21] never mind the issue is not what I thought it was [19:52] Okay so I am still having an issue [19:53] update_relation_data does not seem to be doing its job [19:53] (or I don't understand it) [19:54] https://paste.ubuntu.com/p/7CGKZGmTBc/ [19:54] I am making that call and seeing [19:55] https://paste.ubuntu.com/p/CYzTXVpxCW/ [19:56] lol I see the bug. - not _ [19:56] okay new issue [19:57] the charm reads both the unit and app relation data on relation_changed [19:57] but I can only set one of those at once [19:58] so whichever I set first, I will get a failure saying the other wasn't set [19:59] solved that too [20:09] Issue operator#416 closed: Introspection for events [20:09] PR operator#436 closed: Give ObjectEvents a useful __repr__ [20:14] PR operator#440 opened: Set sys.breakpointhook in main instead of Framework.__init__ [20:15] facubatista: ^ 440 if you're still here [21:26] dstathis, that was fun :) [21:27] Chipaca, will review