[16:12:41] halfak|Mobile_: any plans to make the `features` flag work along with passing multiple rev_ids? [16:14:08] ragesoss: purposefully not supporting that [16:14:39] too expensive? [16:57:42] halfak: I just realized that the switch to https-only had broken my data import. I've updated my app now. Just a heads-up, since you otherwise went out of your way to ensure backwards compatibility after the new API launched. [16:57:58] Sorry about that :( [16:58:44] no worries! I might be the only client that was affected. [17:11:44] hello schana. [17:11:51] I have a very important question for you. :D [17:11:52] hey leila [17:12:01] do you think keeping the backlog grooming today is useful? [17:12:15] Aaron is in the other part of the world, busy with more important things. [17:12:24] I have just got back and have very little tasks. [17:12:52] schana, ^. [17:13:47] yeah, with people out, there's probably not anything that needs addressed that couldn't wait [17:14:08] schana: agreed. I'll reach out to Ellery to check with him. I'll cancel the meeting if he agrees, too. [17:14:08] were you here when we split it into a backlog board and active board? [17:14:09] thanks! [17:14:15] yes, schana. :) [17:50:10] o/ Amir1 [17:53:18] earlier puppet ores::base used to copy the ores-wikimedia-config repo. Today when I was trying fab initialize_server, it was failing with /srv/ores/config already exists. Saw that intialize_server now doing the git clone of ores-wm-config. Because of the the preached role deploy is failing. What are are your thoughts? [17:54:06] Amir1, YuviPanda ^ [18:22:06] sabya: we changed that situation [18:22:45] Amir1: ok, what is the new flow? [18:22:55] we need to make this patch merged: https://gerrit.wikimedia.org/r/280247 [18:23:12] after that the fab would do the initializing [18:23:18] and the puppet doesn't [18:23:45] It would be great if someone with +2 access in puppet merge this [18:24:45] got it [18:26:13] Amir1: I can merge if you want [18:26:25] Amir1: can you check if git::clone is referenced in a require elsewhere? [18:26:34] Amir1, question: so if fab is supposed to get the ores-wm-config, then how do we get the fabfile? [18:26:35] YuviPanda: it would be great [18:27:35] YuviPanda: https://github.com/wiki-ai/ores-wikimedia-config/blob/master/fabfile.py#L82 [18:27:39] we added it here [18:28:19] sabya: good question: I don't know about fabric but I assume we clone it in a central repo [18:28:34] and that copies them to the target [18:29:16] ok. so say I want to launch an "web" instance. what would be the steps? [18:29:16] if we want to download the fabfile and run their commands in the host we are on it, why we don't run it directly? ;) [18:29:37] :) [18:29:39] Amir1: merged [18:29:43] \o/ [18:29:51] thanks YuviPanda :) [18:29:57] np [18:30:02] afk again [18:30:08] thanks [18:30:11] sabya: hmm [18:30:22] you need to go on a host [18:30:42] and fab should connect to the target (mentioned in the fabfile) [18:30:51] *targets [18:31:01] and fab runs the commond [18:31:07] *commands [18:31:18] we are actually moving to scap3 [18:31:41] (that's my job and I'm working on it for about a month) [18:33:49] ok. I think then we need to change the self hosted puppet flow. earlier I could apply the ores::web, puppet::self roles, and run fab initialize. [18:34:03] for testing puppet classes. [18:34:39] you can do it now [18:34:45] since the patch is merged :) [18:35:22] oh ya :) [18:35:37] I don't know how the puppetmasters updates [18:35:46] but probably it's okay now [18:36:27] thanks. I can now move forward. [18:37:06] sorry for being blocked [18:37:38] np :) [20:01:12] I'm not going to join the documentation sprint today, at least not right now; I'm at the annual plan narrative thing.