| [10:31:45] | * freiheit has quit (Quit: Leaving.) |
| [10:57:52] | * mstenta has quit (Quit: Leaving.) |
| [10:59:20] | * bgm has quit (Ping timeout: 244 seconds) |
| [11:03:32] | * bgm has joined #aegir |
| [11:03:45] | * bgm has quit (Changing host) |
| [11:03:45] | * bgm has joined #aegir |
| [11:06:53] | * cweagans has quit (Quit: Gone.) |
| [11:08:10] | * cweagans has joined #aegir |
| [11:21:44] | * cweagans__ has joined #aegir |
| [11:22:20] | * cweagans__ has left #aegir ("Gone.") |
| [11:22:21] | * cweagans__ has joined #aegir |
| [11:22:38] | <cweagans__> | test message (ignore me - testing znc) |
| [11:23:41] | * gandhiano has quit (Ping timeout: 260 seconds) |
| [11:24:22] | * cweagans__ has quit (Client Quit) |
| [11:27:10] | * gusaus has quit (Quit: gusaus) |
| [11:30:18] | * kristofferw has quit (Read error: Connection reset by peer) |
| [12:01:19] | * ivanjaros has joined #aegir |
| [12:10:50] | * cweagans is now known as cweagans_afk |
| [12:21:53] | <formatC^vt> | Add site validation after submit, add/del buttons, variables form validation and rename titles |
| [12:22:01] | <formatC^vt> | and i go to sleep =) |
| [12:22:52] | <formatC^vt> | to cleanup branch |
| [12:34:26] | * Egyptian[Home] has quit (Quit: Leaving.) |
| [12:35:09] | * titanous has quit () |
| [12:53:27] | * gusaus has joined #aegir |
| [13:16:08] | * cweagans_afk is now known as cweagans |
| [13:24:48] | * Yaazkal has quit () |
| [13:50:37] | * gusaus has quit (Quit: gusaus) |
| [13:55:20] | * cweagans is now known as cweagans_afk |
| [14:19:20] | * cweagans_afk is now known as cweagans |
| [14:20:08] | * cweagans is now known as cweagans_afk |
| [15:06:45] | * ivanjaros has quit (Quit: https://drupal.org/user/135190) |
| [15:55:55] | * ivanjaros has joined #aegir |
| [15:56:40] | * David_Hernandez has joined #aegir |
| [15:58:10] | * stijnvbrande has joined #aegir |
| [16:28:18] | * boshtian has joined #aegir |
| [17:05:01] | * Deciphered is now known as DecipheredAFK |
| [17:41:23] | * tobias- has quit (*.net *.split) |
| [17:41:24] | * spyd has quit (*.net *.split) |
| [17:41:24] | * fufroma_ has quit (*.net *.split) |
| [17:41:24] | * orangey has quit (*.net *.split) |
| [17:42:51] | * tobias- has joined #aegir |
| [17:42:51] | * spyd has joined #aegir |
| [17:42:51] | * fufroma_ has joined #aegir |
| [17:42:51] | * orangey has joined #aegir |
| [17:45:50] | * fufroma_ has quit (Ping timeout: 265 seconds) |
| [17:46:17] | * spyd has quit (Ping timeout: 265 seconds) |
| [17:46:48] | * tobias- has quit (Ping timeout: 256 seconds) |
| [17:46:50] | * orangey has quit (Ping timeout: 256 seconds) |
| [17:47:41] | * spyd has joined #aegir |
| [17:48:09] | * tobias- has joined #aegir |
| [17:48:35] | * orangey has joined #aegir |
| [17:52:32] | * fufroma has joined #aegir |
| [17:57:07] | * kristofferw has joined #aegir |
| [18:00:42] | * gandhiano has joined #aegir |
| [18:05:14] | * e-anima has joined #aegir |
| [18:25:13] | * gandhiano has quit (Ping timeout: 244 seconds) |
| [18:35:24] | * thunderWilly has joined #aegir |
| [18:35:45] | * e-anima has quit (Ping timeout: 250 seconds) |
| [19:01:54] | * ratioweb has joined #aegir |
| [20:45:59] | * Egyptian[Home] has joined #aegir |
| [22:01:17] | * mstenta has joined #aegir |
| [22:13:13] | * zombiebeard has joined #aegir |
| [22:15:31] | * zombiebeard has quit (Client Quit) |
| [22:31:48] | * mstenta has quit (Ping timeout: 250 seconds) |
| [22:42:25] | * zombiebeard has joined #aegir |
| [22:54:04] | * zz_drakythe is now known as drakythe |
| [23:08:09] | * mstenta has joined #aegir |
| [23:43:31] | <ergonlogic> | anarcat: ping |
| [23:53:47] | <ergonlogic> | hefring: log pointer |
| [23:53:47] | <hefring> | http://hefring.mig5.net/bot/log/aegir/2015-08-13#T583289 |
| [23:55:58] | * cweagans_afk is now known as cweagans |
| [23:56:04] | <cweagans> | Morning everyone |
| [23:56:25] | <gboudrias> | Morning |
| [23:56:37] | <helmo> | hi |
| [23:56:37] | <hefring> | hey |
| [23:56:38] | <ergonlogic> | hi all |
| [23:56:59] | <ergonlogic> | I was just reviewing last week's scrum: https://hefring.mig5.net/bot/log/aegir/2015-08-06#T581925 |
| [23:57:19] | <formatC^vt> | nihao =) |
| [23:57:31] | <ergonlogic> | cweagans: you setting up an irc proxy? |
| [23:57:39] | * titanous has joined #aegir |
| [23:57:42] | <ergonlogic> | if so, I recommend irssi |
| [23:58:01] | <ergonlogic> | I run it in a detached screen session on a server, and it hasn't failed me yet |
| [23:58:18] | <cweagans> | Yeah, I set up ZNC yesterday. I would have considered other options, but it's the only one that I could find that runs on my NAS. |
| [23:58:24] | <cweagans> | Which arrived yesterday :) |
| [23:59:00] | <ergonlogic> | ok :) |
| [23:59:12] | <ergonlogic> | so, following up on a couple items: |
| [23:59:39] | <ergonlogic> | mailman support signup to mailing lists via the web interface, which helmo linked to in our docs |
| [00:00:05] | <ergonlogic> | or via mailing the list with LISTNAME-join@DOMAIN |
| [00:00:24] | <ergonlogic> | so, for example, discuss-join@aegirproject.org |
| [00:00:34] | <ergonlogic> | we should properly document that |
| [00:00:38] | <ergonlogic> | and tweet about it |
| [00:01:09] | <cweagans> | definitely |
| [00:01:17] | <ergonlogic> | for some reason, I'm finding myself suffering from twitter-aversion |
| [00:01:31] | <ergonlogic> | so if anyone else could take that up, I'd appreciate it |
| [00:01:42] | <ergonlogic> | the password is in lastpass now |
| [00:02:27] | <ergonlogic> | continuing to follow-up on last week's items: |
| [00:02:37] | <cweagans> | helmo: documented mailing list stuff here: https://aegir.readthedocs.org/en/3.x/help/contact/ |
| [00:02:43] | <cweagans> | I can tweet about it later if you want. |
| [00:02:52] | <helmo> | We should also start redirecting more pages from the old community site to the new docs... e.g. also the /contact page |
| [00:03:11] | <ergonlogic> | i didn't get around to releasing Rán, since I can't find a stable combination of D8 core, Features, etc. |
| [00:03:14] | <ergonlogic> | agreed |
| [00:03:42] | <ergonlogic> | I'm just going to post the mess that it is to the aegirproject github account, and apologize in advance for the shabby state of it |
| [00:04:18] | <ergonlogic> | chasing D8 Features is an exercise in frustration |
| [00:04:20] | <ergonlogic> | anyway... |
| [00:04:20] | <cweagans> | ha, that's fine. We can always improve on it. Composer might be a good way to go for managing the specific versions of dependencies. We've been using it for our distro internally and it seems to work pretty well. |
| [00:04:36] | <ergonlogic> | I'd like to see an example of that |
| [00:04:43] | <ergonlogic> | I have a makefile, so that isn't the problem |
| [00:04:59] | <ergonlogic> | it's actually finding the magic combination of commits that allowed it to work |
| [00:05:34] | <ergonlogic> | I'm going to roll back to the date that mpotter did his DCLA demo |
| [00:05:42] | <ergonlogic> | anyway... |
| [00:06:04] | <gboudrias> | Heh, sounds like dev :) |
| [00:06:16] | <ergonlogic> | I didn't get the proposal out to business@ either :-/ |
| [00:07:23] | <helmo> | Where's is the time ... we all seem to need more of it ;) |
| [00:07:45] | <ergonlogic> | cweagans and I have been in touch with mlhess, who will provide a demo in a few weeks of his Kubernetes-backed provisioning of Drupal |
| [00:08:00] | <ergonlogic> | yeah... |
| [00:08:28] | <helmo> | nice to see more of that |
| [00:08:51] | * johnstorey has joined #aegir |
| [00:09:23] | <ergonlogic> | I think we need to reach out to the other folks who'd expressed interest in helping with AegirNG |
| [00:09:35] | <ergonlogic> | or, at least, tracking its progress |
| [00:09:48] | <ergonlogic> | Pronovix, Wunderkraut, etc |
| [00:09:58] | <ergonlogic> | I'll do that by email this week |
| [00:10:36] | <helmo> | maybe one more mailinglist ? |
| [00:11:02] | <ergonlogic> | I think develop@ is probably sufficient for that |
| [00:11:19] | <ergonlogic> | I meant that I'll email them directly to ask if they're still interested |
| [00:11:31] | <cweagans> | +1 for using develop@ |
| [00:11:34] | <ergonlogic> | and direct them to the mailing lists, and our discussions here in IRC |
| [00:12:00] | <ergonlogic> | I actually think we might want to close the user@ mailing list, since it overlaps with discuss@ |
| [00:12:11] | <helmo> | ok, then lets advertise that @develop is the for that ... I'll add a note on the new contact page |
| [00:12:49] | <ergonlogic> | and re-start announce@aegirproject.org, as opposed to being under '@list.aegirproejct.org' |
| [00:13:30] | <gboudrias> | Good ideas for the lists |
| [00:13:39] | <ergonlogic> | cweagans and I have started documenting personas for the re-design of aegirproject.org in hackpad, too |
| [00:13:39] | <helmo> | done |
| [00:14:09] | <ergonlogic> | https://aegir.hackpad.com/Aegirproject.org-Strategy-Fa6e8yEQkMd |
| [00:14:36] | <cweagans> | If you know anyone with web strategy experience that can help us out for an hour or two, that would be really great. |
| [00:14:59] | <cweagans> | I tried to just dive in and start on it myself, but found the process daunting and confusing, so I'm thinking that's better left to somebody that knows what they're doing :) |
| [00:15:45] | <ergonlogic> | I elaborated it a bit, but it'd definitely be improved by someone with more experience in this |
| [00:15:56] | <cweagans> | Once we have the personas defined, we'll move on to the user stories for the website, decide what stories/personas to prioritize, and move forward from there. |
| [00:16:27] | <ergonlogic> | so, in other news, gboudrias has joined the core team |
| [00:16:31] | <ergonlogic> | as y'all know already |
| [00:16:36] | <cweagans> | yaaaaaaaaaaaaaaaaayyyyyyyyyyyyyyy |
| [00:16:38] | <ergonlogic> | welcome gboudrias |
| [00:16:41] | <cweagans> | welcome :) |
| [00:16:49] | <gboudrias> | hahah thanks |
| [00:16:50] | <ergonlogic> | he's posted a 3.1 release issue on d.o |
| [00:17:21] | <helmo> | should have been done earlier |
| [00:17:24] | <ergonlogic> | brb, continue without me, and i'll catch up |
| [00:17:25] | <gboudrias> | yeah, sorry helmo for duplicating the issue |
| [00:17:55] | <helmo> | np, you've created a good list to work from |
| [00:18:09] | <gboudrias> | So in short there's a lot to commit and a lot I would like to review |
| [00:18:40] | <gboudrias> | But the stated plan is to just commit what gets to RTBC before next Thursday |
| [00:18:53] | <helmo> | I did a little review there this week, but unfortunately had little time to spare |
| [00:19:05] | <ergonlogic> | back... part of the challenge is the length of the release process, and so-so tooling |
| [00:19:54] | <helmo> | I'ts getting a bit better, now that we have the release docs in https://aegir.readthedocs.org/en/3.x/community/release-process/ |
| [00:20:11] | <ergonlogic> | if we could optimize any of that, it'd allow for more frequent releases |
| [00:20:12] | <cweagans> | Yeah, we should do this differently in Aegir 4. I think a good target to shoot for would be 15-30 min |
| [00:20:20] | <ergonlogic> | agreed |
| [00:20:21] | <helmo> | one 'blocker' for a release is the debian build environment |
| [00:20:33] | <ergonlogic> | which is where OBS could come in handy |
| [00:20:52] | <cweagans> | What is OBS? |
| [00:20:58] | <ergonlogic> | https://build.opensuse.org/ |
| [00:21:03] | <helmo> | would be really nice ... but we nees someone with specfile experiance |
| [00:21:14] | <ergonlogic> | indeed |
| [00:21:37] | <ergonlogic> | I think the initial target, should we go this way, would be to simply replicate the .deb builds for Debian and Ubuntu |
| [00:22:00] | <gboudrias> | I don't get it, OBS the recording software? |
| [00:22:15] | <gboudrias> | oh the build service |
| [00:22:19] | <helmo> | obs is a service opensuse provides |
| [00:22:19] | <gboudrias> | Makes a lot more sense :p |
| [00:22:42] | <ergonlogic> | but that initial work should make building packages for centos, etc. easier |
| [00:22:45] | <helmo> | you upload an rpm specfile and it builds an rpm |
| [00:22:53] | <helmo> | and as a bonus also debian packages |
| [00:23:21] | <ergonlogic> | ... or maybe the opposite of what I just said :p |
| [00:24:06] | <cweagans> | I'd say just open an issue with some links and whoever has time can take a look |
| [00:24:09] | <ergonlogic> | anyway, it also provide a 'downloads' page that we could presumably cname under dl.aegirproject.org or whatever |
| [00:24:44] | <cweagans> | or if you want, I can do that |
| [00:24:46] | <ergonlogic> | it'd start to simplify some o the less well documented/maintained jenkins jobs |
| [00:25:10] | <helmo> | we already have an issue for that https://www.drupal.org/node/2281423 |
| [00:25:11] | <hefring> | https://www.drupal.org/node/2281423 => Make RPMs to add support for RHEL/CentOS [#2281423] => 6 comments, 2 IRC mentions |
| [00:25:16] | <ergonlogic> | cweagans: I'd really like to hear more about any openshift/kubernetes progress you've made |
| [00:25:36] | <cweagans> | we'd presumably still need our own Jenkins server to do the integration testing for commits, but if that's all we have, we're in pretty good shape. |
| [00:25:45] | <ergonlogic> | also, a potential new university user has tentatively offered to help with RPMs |
| [00:25:57] | <cweagans> | so re openshift/kubernetes, I was able to get openshift up and running locally and the interface/ux is, to put it bluntly, horrible. |
| [00:26:12] | <ergonlogic> | right, re ci on jenkins |
| [00:26:38] | <cweagans> | I think it's meant to be more of a status indicator and all the real heavy lifting is done through either the Openshift API or the cli tool (which talks to the same api) |
| [00:27:39] | <cweagans> | In any case, I've been trying to think through the architecture a bit, and the Drupal site can actually be pretty stateless if we want to go this route. We wouldn't even need a separate db container to make it work. |
| [00:28:39] | <cweagans> | D8 has the concept of entity storage, so we can create an "entity" that's basically the output of some openshift api endpoint (i.e. give me a list of all projects running on openshift) |
| [00:29:02] | <helmo> | brb |
| [00:29:04] | <cweagans> | and we can do the same for lots of other things in the openshift environment, including users and their associated roles and permissions |
| [00:29:30] | <cweagans> | So deploying Aegir becomes, essentially, 1) deploy openshift and 2) spin up the Aegir container on top of it. |
| [00:29:30] | <ergonlogic> | there's a 'remote_entities' project, iirc, that could provide us the entity hooks and other integration points with Drupal |
| [00:29:44] | <cweagans> | Yeah, that's kind of a shim to make the same sort of thing work in D7 |
| [00:29:57] | <cweagans> | but there's an actual core api for it in d8. We could port the module to make the process nicer, though |
| [00:30:02] | <cweagans> | https://www.drupal.org/project/remote_entity |
| [00:30:50] | <cweagans> | The other thing with doing it this way, though, is that we're more or less tied to Openshift. |
| [00:30:57] | * fatguylaughing has joined #aegir |
| [00:31:23] | <ergonlogic> | yeah, I think we'd be ok with that for Aegir4 |
| [00:31:39] | <ergonlogic> | but it'd be really nice to support other tools in Aegir5+ |
| [00:31:53] | <ergonlogic> | in the spirit of 'tools not policy' |
| [00:33:05] | <cweagans> | I mean, from a code standpoint, it might be entirely possible to swap out the OpenshiftProvisioner service (or whatever the class will be called) for...idk...the HerokuProvisioner or whatever, but the main thing is that we wouldn't be thinking through the integration points for those other PaaSes, and that's really where things get murky. |
| [00:33:28] | <cweagans> | I'm pretty okay with a hard dependency on openshift though |
| [00:33:58] | <cweagans> | we want to prescribe best practices, and these seem like the best tools available, so I'm happy if you're happy. |
| [00:34:11] | <ergonlogic> | I think we need to get really familiar with how a full stack container orchestration engine works |
| [00:34:25] | <ergonlogic> | at this point, I think openshift is the only libre alternative |
| [00:34:36] | <cweagans> | Yeah, for sure. It doesn't help that they all do it differently, though. |
| [00:34:46] | <ergonlogic> | as flynn catches up, we should consider it too |
| [00:35:03] | <ergonlogic> | right, but I suspect there's a level of abstraction we could perhaps layer on top |
| [00:35:13] | <helmo> | back |
| [00:35:46] | <ergonlogic> | right now, I feel like there are too many unknowns |
| [00:35:51] | <cweagans> | Yeah, I agree. We just have to make sure not to paint ourselves into a corner and get the lowest common denominator of all the PaaSes that we claim to support. |
| [00:35:52] | <cweagans> | agreed |
| [00:35:59] | <cweagans> | there's so much to learn about this stuff :( |
| [00:36:16] | * fatguylaughing has quit (Quit: fatguylaughing) |
| [00:36:19] | <ergonlogic> | right, well, let's focus on openshift and get a handle on that |
| [00:36:57] | <ergonlogic> | the next version after that can target other PaaSes |
| [00:37:00] | <cweagans> | Cool |
| [00:37:10] | <ergonlogic> | assuming some have caught up with openshift by then |
| [00:37:28] | <cweagans> | So for anyone interested (and I already mentioned this a couple days ago), there's a good video series that serves as an intro to openshift: https://www.youtube.com/watch?v=b5THfQzVUqk |
| [00:37:43] | <ergonlogic> | I still think some level of abstraction in Aegir will be req'd |
| [00:37:51] | <cweagans> | It says part 1 of 3, but there are like 10-15 parts, I think |
| [00:38:08] | <ergonlogic> | otherwise, we're asking users to get too familiar w/ openshift's internals, imo |
| [00:38:27] | <ergonlogic> | yeah, I'm going to prioritize making time to watch those |
| [00:38:39] | <ergonlogic> | might be good to list them in the 4.x branch of our docs |
| [00:38:50] | <cweagans> | Yeah, I think we should define the content model for the frontend, and then figure out how to translate our concepts (Projects, Environments, Services, etc) to Openshift concepts |
| [00:38:57] | <ergonlogic> | and centralize our documentation efforts there, via-à-vis aegirng |
| [00:39:52] | <ergonlogic> | right |
| [00:40:08] | * fatguylaughing has joined #aegir |
| [00:42:25] | <ergonlogic> | I think targetting jekyll as a first supported app should allow us to get to a work proof-of-concept reasonably soon |
| [00:42:45] | <cweagans> | +1 |
| [00:43:06] | <ergonlogic> | and allow us to self-host our new site, which'll address some community members' concerns ;) |
| [00:43:14] | <cweagans> | ha, for sure |
| [00:43:27] | <ergonlogic> | it avoids all the hard stuff |
| [00:43:33] | <ergonlogic> | unless i'm mistaken |
| [00:43:41] | <cweagans> | No, that's accurate. |
| [00:43:49] | <cweagans> | site with no services is the easy part. |
| [00:44:05] | <ergonlogic> | right, just an nginx container or whatever with static html |
| [00:44:24] | <cweagans> | So going forward, I'm going to re-watch those openshift videos and then I'm going to focus on non-technical, organizational stuff for the project, mainly the website and the conservancy application. If everything goes as expected, the initial documentation required for the website design to move forward and the conservancy application should be either done or very close to done by the next time we talk. |
| [00:44:26] | <cweagans> | exactly. |
| [00:44:43] | <ergonlogic> | ok, nice |
| [00:45:18] | <ergonlogic> | fwiw, I seem to recall jen@backdrop mentioning that they've been on the SF conservancy waiting list for 8 months |
| [00:45:25] | <cweagans> | pm? |
| [00:45:39] | <ergonlogic> | so let's not expect any quick responses there, or depend on ti |
| [00:45:40] | <ergonlogic> | it |
| [00:46:41] | * boshtian has quit (Quit: boshtian) |
| [00:48:12] | <cweagans> | Okay, any other updates from the core team? |
| [00:48:18] | <ergonlogic> | we should still try to get our application in asap |
| [00:48:28] | <ergonlogic> | since the waiting period doesn't start until then |
| [00:48:31] | <cweagans> | Right |
| [00:51:07] | <cweagans> | Alright, I think that's a wrap! Thanks everyone. |
| [00:51:12] | <cweagans> | hefring: log pointer? |
| [00:51:12] | <hefring> | http://hefring.mig5.net/bot/log/aegir/2015-08-13#T583468 |
| [00:51:47] | <helmo> | thanks, see you all next week... |
| [00:52:29] | <cweagans> | ergonlogic: which list do you send the scrum log to? core? develop? |
| [00:52:38] | <cweagans> | (both?) |
| [00:52:58] | <gboudrias> | :) |
| [00:53:29] | <ergonlogic> | umm, I haven't been |
| [00:53:37] | <ergonlogic> | but develop would probably be best |
| [00:53:49] | <ergonlogic> | I used to post them to the community page |
| [00:54:06] | <ergonlogic> | I mean site |
| [00:54:38] | * hestenet has joined #aegir |
| [00:54:39] | <ergonlogic> | core is private, and there's really no reason not to keep these discussions public |
| [00:55:19] | * cweagans is now known as cweagans_afk |
| [00:55:39] | * cweagans_afk is now known as cweagans |
| [00:55:43] | <cweagans> | okay, deal |
| [00:59:47] | <cweagans> | sent |
| [00:59:55] | * fatguylaughing has quit (Quit: fatguylaughing) |
| [01:00:23] | * mstenta has quit (Quit: Leaving.) |
| [01:02:10] | * mstenta has joined #aegir |
| [01:22:51] | * David_Hernandez has quit (Quit: :wq!) |
| [01:25:29] | <gboudrias> | helmo: Can I PM you for a contact question? |
| [01:40:04] | * Yaazkal has joined #aegir |
| [01:41:14] | <gboudrias> | Nvm, see email |
| [01:49:27] | * Yaazkal has quit () |
| [01:49:44] | * freiheit has joined #aegir |
| [01:53:06] | <cweagans> | mlhess: Are you subscribed to the develop list ? |
| [01:55:24] | * cweagans is now known as cweagans_afk |
| [01:55:39] | <gboudrias> | gboudrias: bgm I don't know if you're on the develop list but we just published https://www.drupal.org/project/hosting_variables |
| [01:56:13] | <gboudrias> | I'm working towards a SaaS utils module that will depend on it |
| [01:56:19] | <bgm> | cool |
| [01:56:30] | <bgm> | how do we subscribe to the dev list? |
| [01:56:50] | <gboudrias> | bgm: https://listes.koumbit.net/cgi-bin/mailman/listinfo/develop-aegirproject... |
| [01:56:55] | * cweagans_afk is now known as cweagans |
| [01:57:08] | <gboudrias> | Yeah it seems we're all setting a user email and whatnot, so that's what saas utils will do |
| [01:58:27] | <cweagans> | bgm gboudrias: or just send an email to develop-join@aegirproject.org |
| [01:58:30] | <bgm> | or disabling those annoying security upgrade messages ;) |
| [01:58:34] | <gboudrias> | right |
| [01:58:43] | <gboudrias> | hah, yeah |
| [02:01:56] | <cweagans> | mlhess: if you're not subscribed, you might want to take a look at https://listes.koumbit.net/pipermail/develop-aegirproject.org/2015-Augus..., as it could be useful to you :) |
| [02:07:54] | * stijnvbrande has quit (Quit: Connection closed for inactivity) |
| [02:13:27] | * hestenet has quit (Remote host closed the connection) |
| [02:15:59] | * jackinloadup has joined #aegir |
| [02:16:02] | * hestenet has joined #aegir |
| [02:23:59] | * fatguylaughing has joined #aegir |
| [02:24:40] | * ivanjaros has quit (Quit: https://drupal.org/user/135190) |
| [02:33:44] | * hestenet has quit (Read error: Connection reset by peer) |
| [02:34:18] | * hestenet has joined #aegir |
| [02:40:36] | * fatguylaughing has quit (Quit: fatguylaughing) |
| [02:44:01] | * hestenet has quit (Remote host closed the connection) |
| [02:44:26] | * hestenet has joined #aegir |
| [02:45:46] | * ivanjaros has joined #aegir |
| [02:57:49] | * Yaazkal has joined #aegir |
| [03:03:07] | * hestenet has quit (Remote host closed the connection) |
| [03:09:56] | * zombiebeard has quit (Quit: zombiebeard) |
| [03:14:15] | * hestenet has joined #aegir |
| [03:25:21] | <ergonlogic> | freiheit: so, we're good for 11am PST? |
| [03:26:01] | <ergonlogic> | if so, I'll just grab a quick lunch and ping you in 35 mins |
| [03:36:51] | <gboudrias> | ergonlogic: Re https://www.drupal.org/node/2547209 , what's a good message? "This task didn't necessarily take an entire second to execute"? |
| [03:36:51] | <hefring> | https://www.drupal.org/node/2547209 => Display execution time in task log [#2547209] => 9 comments, 1 IRC mention |
| [03:37:08] | * hestenet has quit (Remote host closed the connection) |
| [03:37:47] | * hestenet has joined #aegir |
| [03:43:00] | * gusaus has joined #aegir |
| [03:48:21] | <ergonlogic> | hmm |
| [03:55:36] | <ergonlogic> | gboudrias: "Many tasks take less than 1 second to perform. This task represents an aggregate of the preceding tasks' duration." ? |
| [03:56:11] | <gboudrias> | ergonlogic: Sounds good |
| [03:56:50] | * gusaus has quit (Ping timeout: 240 seconds) |
| [03:57:49] | * gusaus has joined #aegir |
| [03:58:04] | * cweagans has left #aegir ("Textual IRC Client: www.textualapp.com") |
| [03:58:07] | * cweagans has joined #aegir |
| [03:59:07] | <cweagans> | ergonlogic: I was thinking through task execution process for Openshift operations. I think the big thing that we need to test is whether or not Openshift/Kubernetes have the capability to maintain an internal queue of work that needs done. If not, we might need to have a queue after all :/ |
| [03:59:30] | <ergonlogic> | ok |
| [03:59:36] | <cweagans> | that is, if I submit 100 jobs to Openshift all at once, will it try to simultaneously execute them all? Or will it do them in the order it receives them? |
| [03:59:47] | <ergonlogic> | I'm not really worried about that too much |
| [04:00:04] | <ergonlogic> | Celery appears feature complete, afaict |
| [04:00:14] | <ergonlogic> | and supports scheduled tasks, etc. |
| [04:00:22] | * e-anima has joined #aegir |
| [04:00:44] | <ergonlogic> | while it puts the burden on us, at least there's some further value-add that we can provide |
| [04:00:44] | <cweagans> | True, but it dictates whether or not we go in that direction or not. Nightly site backups are a case where something like that can happen regularly. |
| [04:01:30] | * thunderWilly has quit (Ping timeout: 240 seconds) |
| [04:02:23] | <ergonlogic> | right, I just mean that the implentation shouldn't require much work. If it's needed, we should be able to focus on the business logic |
| [04:06:11] | * fatguylaughing has joined #aegir |
| [04:13:45] | * zombiebeard has joined #aegir |
| [04:19:32] | <cweagans> | fair enough |
| [04:24:30] | * ratioweb has quit (Ping timeout: 260 seconds) |
| [04:33:53] | * gandhiano has joined #aegir |
| [04:44:36] | <ergonlogic> | sorry, I'm enthusiastic about Celery largely because of how much time I spent in the weeds with our 'ghetto' queue |
| [04:45:08] | <ergonlogic> | and realizing just how limited it is, and how hard it'd be to build something really good with that as a base |
| [05:01:14] | <cweagans> | For sure. I don't much care about the implementation of the queue just as long as it's not what we have now :P RabbitMQ seems to be popular these days and there's a PHP library to push things into it, but like I said, I really have no strong opinion on the matter. |
| [05:01:24] | <cweagans> | (if we need a queue at all, that is) |
| [05:04:30] | * Yaazkal has quit () |
| [05:14:06] | <ergonlogic> | sure, rabbitmq is the default broker for celery: http://celery.readthedocs.org/en/latest/getting-started/brokers/rabbitmq... |
| [05:14:28] | <ergonlogic> | celery, mind you, is a full-feature *task* queue |
| [05:14:48] | <ergonlogic> | as opposed to just a message queue |
| [05:14:52] | <ergonlogic> | "just" |
| [05:16:51] | <gboudrias> | Anyone wants to review this? https://www.drupal.org/node/2541758 |
| [05:16:53] | <hefring> | https://www.drupal.org/node/2541758 => Trigger verify after "git pull" task when a site is on remote server [#2541758] => 16 comments, 2 IRC mentions |
| [05:16:57] | <gboudrias> | I haven't worked with hosting_git recently |
| [05:17:50] | * gandhiano has quit (Ping timeout: 250 seconds) |
| [05:28:18] | * hestenet has quit (Remote host closed the connection) |
| [05:50:16] | * ivanjaros has quit (Quit: https://drupal.org/user/135190) |
| [05:51:33] | * hestenet has joined #aegir |
| [05:58:54] | <formatC^vt> | gboudrias> i think you are right about wrong place https://www.drupal.org/node/2040445 should do i file a different issue? |
| [05:58:54] | <hefring> | https://www.drupal.org/node/2040445 => Provide a safe way to remove a server from the system [#2040445] => 23 comments, 1 IRC mention |
| [06:01:31] | <gboudrias> | formatC^vt: Not sure |
| [06:02:10] | <gboudrias> | cweagans: https://www.drupal.org/node/2040445 I think the patch doesn't address the larger issue, it seems like there will still be work to do after we commit it |
| [06:02:12] | <hefring> | https://www.drupal.org/node/2040445 => Provide a safe way to remove a server from the system [#2040445] => 23 comments, 2 IRC mentions |
| [06:02:35] | <gboudrias> | Meaning we can't mark it as fixed afaict |
| [06:04:01] | <formatC^vt> | yes, my bad, this issue still needs work |
| [06:04:18] | <gboudrias> | formatC^vt: Well, it's just a matter of bureaucracy at this point :) |
| [06:05:10] | <gboudrias> | ergonlogic: ^ There's a commit without a comment in the issue, did you address the main point? |
| [06:06:38] | <ergonlogic> | which one? the safe server removal? |
| [06:07:12] | <gboudrias> | yes |
| [06:09:31] | <ergonlogic> | I was working on it |
| [06:09:34] | <ergonlogic> | yes |
| [06:09:57] | <gboudrias> | Alright so I assume we shouldn't close it just yet? |
| [06:10:51] | * cweagans is now known as cweagans_afk |
| [06:11:48] | <ergonlogic> | no, close it if it's been resolved |
| [06:11:55] | <ergonlogic> | I committed those changes to 3.x |
| [06:12:15] | <ergonlogic> | so, presumably the recent patch just finished the job I started |
| [06:12:22] | <ergonlogic> | but I haven't looked really closely |
| [06:12:26] | <gboudrias> | Ok great, that's what I wasn't wondering :) |
| [06:12:29] | <gboudrias> | was** |
| [06:12:53] | <gboudrias> | formatC^vt: We're good, I'll just commit your fix before the release |
| [06:13:19] | <formatC^vt> | ok, thx =) |
| [06:14:25] | * cweagans_afk is now known as cweagans |
| [06:46:30] | * gusaus has quit (Quit: gusaus) |
| [06:50:47] | * hestenet has quit (Read error: Connection reset by peer) |
| [06:51:22] | * azend has joined #aegir |
| [06:51:26] | * hestenet has joined #aegir |
| [06:51:27] | <azend> | hello :) |
| [06:52:20] | <azend> | How does one migrate an aegir2 platform to aegir3? |
| [06:52:28] | * gusaus has joined #aegir |
| [06:52:30] | <gboudrias> | azend: Hi! |
| [06:52:43] | <azend> | It turns out I'm an idiot and I never checked the production aegir version |
| [06:52:56] | <azend> | after going through two vagrant vm |
| [06:52:56] | <azend> | s |
| [06:52:59] | <gboudrias> | Hm? |
| [06:53:01] | <azend> | hi gboudrias :) |
| [06:53:53] | <azend> | I really don't want to start from scratch again |
| [06:53:54] | <gboudrias> | azend: Well, you mean migrate all sites from a Aegir2 sever to an Aegir3 server? |
| [06:54:14] | <azend> | sure |
| [06:54:43] | <gboudrias> | azend: You can use hosting_remote_import , you can test this patch :) https://www.drupal.org/node/2538788 |
| [06:54:44] | <hefring> | https://www.drupal.org/node/2538788 => Make this module compatible with Aegir2 [#2538788] => 0 comments, 1 IRC mention |
| [06:55:16] | <gboudrias> | azend: For remote import to work, you'll need to first re-create the platform, usually with a makefile |
| [06:55:38] | <gboudrias> | (Sorry if this is too obvious, I don't know how much you much) |
| [06:55:56] | <azend> | I'm pretty brand new to aegir |
| [06:56:05] | <gboudrias> | Okay, well usually you'll have created the platform with a makefile, so you can just use the same one |
| [06:56:26] | <azend> | but I thought the workflow was basically drush provision-backup |
| [06:56:34] | <azend> | copy the backup to another machine |
| [06:56:51] | <gboudrias> | If this is not the case, you can just copy the platform over and empty the sites folder |
| [06:56:58] | <azend> | expand the backup on the other machine in /var/aegir/platforms |
| [06:57:08] | <azend> | copy over nginx/apache2 configs |
| [06:57:13] | <azend> | and copy the dbs too |
| [06:57:27] | <azend> | and then add the platform in aegir frontend |
| [06:57:32] | <gboudrias> | azend: I don't think provision-backup works on platforms, at least I've never used it in that way |
| [06:58:03] | <gboudrias> | There is a suggested workflow for mass import but I've never seen it work, I think you might have a much easier time with remote_import |
| [06:58:28] | <azend> | gboudrias: maybe |
| [06:58:42] | <azend> | aegir has so much easily findable incorrect information on google |
| [06:58:45] | <gboudrias> | The distinction between sites and platforms is pretty crucial, since you want to copy the platform's codebase but you'll need to move the DBs too for the sites to work |
| [06:58:49] | <azend> | not because it's bad but because it's out of date |
| [06:58:57] | <gboudrias> | Yeah :/ |
| [06:59:24] | <gboudrias> | azend: If you just edit the drushrc files manually they'll stop working once Aegir rewrites them |
| [06:59:33] | <azend> | gboudrias: I didn't know remote import existed :) |
| [07:00:10] | <cweagans> | azend: wait, you're trying to bring up aegir in a vm? |
| [07:00:18] | <cweagans> | azend: there's a thing for that. |
| [07:00:18] | <azend> | yeah |
| [07:00:19] | <cweagans> | sec |
| [07:00:39] | <azend> | cweagans: cool |
| [07:00:41] | <cweagans> | azend: https://github.com/GetValkyrie/valkyrie |
| [07:00:44] | <azend> | I've done it 3 times now |
| [07:00:54] | <azend> | just the downloading and moving files around is slow |
| [07:02:06] | <azend> | cweagans: interesting |
| [07:02:16] | <gboudrias> | Oh I thought you were just testing for prod, my bad. Valkyrie is great. |
| [07:02:32] | <cweagans> | azend: That was built primarily by ergonlogic, the project lead. It's great. |
| [07:02:55] | <azend> | copying prod to a local dev machine gboudrias |
| [07:03:21] | <gboudrias> | azend: Yeah Valkyrie is pretty much made for that |
| [07:03:57] | <gboudrias> | (Or at least for local development) |
| [07:05:59] | <azend> | I'm using Aegir 2.4 in prod |
| [07:06:04] | <azend> | Does this work with that? |
| [07:06:26] | <azend> | and is there then some helper to allow me to copy over the sites/platforms? |
| [07:07:07] | <azend> | or is that just wishful thinking :) |
| [07:09:34] | <cweagans> | remote_import ought to do the trick |
| [07:09:39] | <cweagans> | I *think* it works across versions |
| [07:09:41] | <cweagans> | but I haven't tried it |
| [07:14:45] | <gboudrias> | afaik you need the patch but it could use more testers |
| [07:15:06] | <ergonlogic> | the challenge you'll have with remote_import (apart from some possible bugs in having Drush5 talk to Drush6, and provision2 talk to provision3, etc) is that you'd need to grant ssh access to the 'aegir' user on your local machine to the remote server |
| [07:15:53] | <ergonlogic> | but, if you're building platforms via makefiles, you can also do the process manually pretty easily |
| [07:16:50] | <ergonlogic> | create an identical platform locally (via drush make), make a backup of the prod site, scp/rsync it down to the local vm, and deploy it |
| [07:17:07] | <ergonlogic> | those last few steps are really all that remote_import does anyway |
| [07:17:56] | * zombiebeard has quit (Quit: zombiebeard) |
| [07:18:50] | <ergonlogic> | azend: Valkyrie is really built around git/Features-based workflows though, so no need to move files and dbs around in most cases |
| [07:19:18] | <ergonlogic> | I'll have to work on getting sync tasks set up properly, which'd simplify that last bit |
| [07:21:07] | * zombiebeard has joined #aegir |
| [07:24:52] | <azend> | cweagans: already patching valkyrie :) |
| [07:24:52] | <azend> | https://github.com/GetValkyrie/valkyrie/pull/72 |
| [07:25:58] | <cweagans> | azend++ |
| [07:26:35] | <azend> | cweagans: are you personally involved with valkyrie? |
| [07:26:48] | <cweagans> | nope, but I'm an aegir maintainer. |
| [07:26:54] | <azend> | wait a second... |
| [07:26:59] | <azend> | [17:02:32] <cweagans> azend: That was built primarily by ergonlogic, the project lead. It's great. |
| [07:27:06] | <azend> | haha |
| [07:27:22] | <azend> | I never put together that ergonlogic was in here talking to us at the same time |
| [07:27:28] | <azend> | now I feel like a fool |
| [07:29:10] | <azend> | woo! patched merged into master |
| [07:29:28] | <azend> | now my valkyrie installation is out of date |
| [07:29:37] | <azend> | ergonlogic++ |
| [07:34:20] | <gboudrias> | ergonlogic: For the drupalcamp session, I'm guessing most Drupal folks in Montreal are aware that Aegir exists, so maybe I should focus more on development? |
| [07:35:34] | <gboudrias> | On top of the general new stuff of course |
| [07:35:55] | * titanous has quit () |
| [07:36:16] | <ergonlogic> | gboudrias: it's tough to say... there always seems to be a new crowd every time we present |
| [07:36:22] | <ergonlogic> | except for a couple regulars |
| [07:36:31] | <azend> | Canadians represent! |
| [07:36:45] | <ergonlogic> | Montreal definitely tends to get more Aegir than most |
| [07:36:55] | <ergonlogic> | azend: I see you're in Kitchener, is it? |
| [07:37:07] | <gboudrias> | Alright, I'll do a bit of everything I suppose |
| [07:37:19] | <azend> | ergonlogic: not exactly |
| [07:37:38] | <azend> | ergonlogic: how'd you figure that one out? |
| [07:37:57] | <ergonlogic> | well, I usually start with something high-level for the newcomers to get their bearings, then dive deeper into whatever newer stuff I've been working on |
| [07:38:03] | <ergonlogic> | azend: your CV |
| [07:38:11] | <azend> | ah |
| [07:38:19] | <azend> | linkedin or azend.org? |
| [07:38:43] | <ergonlogic> | gboudrias: but you're a maintainer now too. so you're approach is as valid as mine :) |
| [07:38:50] | <ergonlogic> | azend.org |
| [07:39:00] | <azend> | :) |
| [07:39:11] | <gboudrias> | hahah |
| [07:39:25] | <azend> | That's such a crusty old site haha |
| [07:39:55] | <ergonlogic> | anything that wasn't deployed in the last 5 minutes is crusty these days |
| [07:40:13] | <gboudrias> | azend: I like it, simple is good :) |
| [07:40:37] | <azend> | but it's simple and static which means it gets served up quick, can be easily cached, and is pretty much invulnerable to attach |
| [07:40:39] | <azend> | attack |
| [07:40:51] | <ergonlogic> | yep |
| [07:42:25] | <ergonlogic> | we're targeting jekyll as our prototype of AegirNG |
| [07:42:43] | <ergonlogic> | ugh |
| [07:42:48] | <ergonlogic> | "in our" |
| [07:43:23] | <ergonlogic> | as a baseline |
| [07:43:45] | * titanous has joined #aegir |
| [07:44:05] | <azend> | ergonlogic: aegirng? |
| [07:44:07] | <ergonlogic> | then tackle the complexities of shared filesystems, db container, etc |
| [07:44:21] | <azend> | mmm ergonlogic you use trello too? |
| [07:44:30] | <ergonlogic> | we've just released Aegir3, and will supporting it for 2-3 years |
| [07:44:45] | <ergonlogic> | but we're architecting a new system |
| [07:45:04] | <ergonlogic> | that'll be a ground-up re-write on top of kubernetes/openshift |
| [07:45:07] | <azend> | ah |
| [07:45:17] | <azend> | nice |
| [07:45:43] | <ergonlogic> | and should eventually support pluggable PaaS backends |
| [07:45:59] | <ergonlogic> | as other PaaSs catch up to openshift |
| [07:46:17] | <azend> | heh catch up |
| [07:46:23] | <azend> | I'm not sure that'll ever happen |
| [07:46:37] | <ergonlogic> | well, catch up to where openshift is today, more or less |
| [07:46:52] | <ergonlogic> | that is, mostly feature complete |
| [07:47:30] | <ergonlogic> | but we aren't targeting that level of flexibility just yet |
| [07:48:18] | <ergonlogic> | first we'll switch to container-based deployments, figure out all their vagaries, an how to abstract it sufficiently |
| [07:48:39] | <ergonlogic> | then we can revisit how to handle multiple backends |
| [07:48:53] | <azend> | ergonlogic: does valkyrie work with drush 8.x? |
| [07:49:35] | <ergonlogic> | not sure |
| [07:49:46] | <azend> | well I guess there's only one way to find out |
| [07:49:47] | <azend> | ;) |
| [07:50:01] | <ergonlogic> | obviously post an issue if not |
| [07:50:18] | <ergonlogic> | for a long time we were chasing head in the Drush 7.x dev cycle |
| [07:50:52] | <ergonlogic> | but the other Drush maintainers are starting to take it in a different direction |
| [07:51:48] | <ergonlogic> | also, they play pretty fast and loose with API changes and such |
| [07:51:58] | <ergonlogic> | so I don't try too hard to keep up anymore |
| [07:52:14] | <ergonlogic> | more trouble than it's worth, imho |
| [07:53:45] | <azend> | ergonlogic: I think not |
| [07:53:45] | <azend> | http://pastie.org/private/qhrtewotd3sucf0ggvwvg |
| [07:54:43] | <ergonlogic> | what's `drush vgp`? |
| [07:55:07] | <azend> | drush vagrant-generate-platform |
| [07:55:21] | <ergonlogic> | you mean vpg? |
| [07:55:24] | <azend> | err valkyrie |
| [07:55:36] | <azend> | ... |
| [07:55:44] | <azend> | sure... |
| [07:56:11] | <azend> | ergonlogic: ;) |
| [07:56:47] | <ergonlogic> | it's subject-verb, instead of verb-subject |
| [07:56:55] | <ergonlogic> | which was probably a bad idea |
| [07:57:14] | <ergonlogic> | sorry object-verb |
| [07:58:34] | <azend> | nah, it's just my fault |
| [07:58:36] | <ergonlogic> | but I wanted the site and platform commands grouped together for auto-completion |
| [07:58:46] | <azend> | consistency is better anyway |
| [07:59:33] | <ergonlogic> | I figure that I'm more likely to know I want to operate on a site, and have to figure out the operation name, than vice-versa |
| [07:59:58] | <azend> | I really like the idea of drush but I wish it were more in parallel with other gnu/*nix tools |
| [07:59:59] | <ergonlogic> | well, it's probably not consistent with most comparable tools |
| [08:00:18] | <ergonlogic> | how so? |
| [08:00:34] | <azend> | gnu tools rarely use = |
| [08:00:45] | <azend> | the commands are all really long |
| [08:00:59] | <azend> | things don't finish immediately |
| [08:01:05] | <azend> | logging isn't at all similar |
| [08:01:17] | <ergonlogic> | true |
| [08:01:28] | <azend> | drush doesn't use stdio properly |
| [08:01:37] | <ergonlogic> | nor stderr |
| [08:02:05] | <ergonlogic> | stdio... right, read that wrong |
| [08:02:20] | <ergonlogic> | yeah, it's a bit of a mess |
| [08:02:24] | <ergonlogic> | more than a bit really |
| [08:02:38] | <ergonlogic> | but that's what you get when a bunch of web devs build a cli tool |
| [08:02:50] | <azend> | you can't easily export a whatever, pipe it through tar, send it through netcat, recieve it on the other end with curl, untar it on the other side replacing all occurrences of a change in domain |
| [08:02:55] | <ergonlogic> | at least that's what you got 8 years ago, in this case |
| [08:03:03] | <azend> | like gnu/*nix users love doing every day |
| [08:03:11] | <ergonlogic> | right |
| [08:03:39] | <ergonlogic> | one of the things I'm trying to bring into Aegir is a more unixy approach |
| [08:03:47] | <ergonlogic> | do one thing and do it well, etc |
| [08:03:50] | <azend> | yeah |
| [08:04:13] | <ergonlogic> | drupal (and by extension drush) tend to take a more monolithic approach |
| [08:04:19] | <ergonlogic> | or have in the past, anyway |
| [08:04:35] | * zombiebeard has quit (Quit: zombiebeard) |
| [08:04:44] | <azend> | drush is fine by itself but has a different feel than other common tools |
| [08:05:01] | <azend> | and drush would get super powers if it played with others better |
| [08:05:20] | <azend> | ergonlogic: nice |
| [08:05:25] | * fatguylaughing has quit (Quit: fatguylaughing) |
| [08:05:25] | <ergonlogic> | Drupal has lived on an island of its own for many years |
| [08:05:42] | <ergonlogic> | and has only started to play nice with others relatively recently |
| [08:06:02] | <ergonlogic> | there was a strong 'not invented here' approach |
| [08:07:18] | <ergonlogic> | even the testing framework in core (while still called simpletest) was pretty much re-implemented, if I'm not mistaken |
| [08:08:39] | <ergonlogic> | ok, gotta go. Nice chatting :) |
| [08:08:45] | <azend> | yeah, thanks |
| [08:08:48] | <azend> | you too |
| [08:19:18] | * gusaus has quit (Quit: gusaus) |
| [08:19:33] | * Egyptian[Home] has quit (Ping timeout: 246 seconds) |
| [08:24:38] | * e-anima has quit (Read error: Connection reset by peer) |
| [08:27:08] | * gusaus has joined #aegir |
| [08:32:05] | * Egyptian[Home] has joined #aegir |
| [08:36:50] | * gusaus has quit (Quit: gusaus) |
| [08:41:40] | * jackinloadup has quit (Read error: Connection reset by peer) |
| [08:54:45] | * hestenet has quit (Remote host closed the connection) |
| [09:00:58] | * formatC_vt has joined #aegir |
| [09:00:58] | * formatC_vt has quit (Changing host) |
| [09:00:58] | * formatC_vt has joined #aegir |
| [09:02:56] | * drakythe is now known as zz_drakythe |
| [09:04:35] | * formatC^vt has quit (Ping timeout: 244 seconds) |
| [09:17:03] | * formatC^vt has joined #aegir |
| [09:20:36] | * formatC_vt has quit (Ping timeout: 244 seconds) |
| [09:38:38] | * cweagans is now known as cweagans_afk |
