rick_h | morning | 11:07 |
---|---|---|
rick_h | anyone have links to the new UX I can peek at. I got a link in a hanguot from sinzui the other day but I can't find it in the juju-gui g-docs folder now :/ | 11:20 |
jcastro | hey guys, I need a relatively complex screenshot of something in the GUI | 12:57 |
jcastro | anyone have anything handy? | 12:57 |
rick_h | jcastro: complex in the form of a big deploy in the environment? | 12:59 |
jcastro | yeah | 13:00 |
* rick_h tries to look up how to load the gui with the old large rapi json setup | 13:01 | |
jcastro | yeah! something hairy. | 13:01 |
rick_h | whoa, takes a little bit to load up the rapi on this, not sure I ever ran this before lol | 13:03 |
rick_h | hmm, not that crazy after all. Just a lot of notes. | 13:03 |
rick_h | jcastro: http://uploads.mitechie.com/lp/gui-large-json.png is what the large.json looks like | 13:05 |
rick_h | jcastro: not sure if you need something w/ w/o the browser and such | 13:05 |
jcastro | dude that is perfect | 13:06 |
rick_h | k, then there you go | 13:06 |
jcastro | rick_h: it'll be in my blog post today! | 13:53 |
rick_h | jcastro: is that ok to blog the browser? | 13:54 |
jcastro | I cropped | 13:54 |
jcastro | only did the blocks/canvas | 13:54 |
rick_h | gotcha, cool | 13:54 |
jcastro | yeah the thought crossed my mind while I was drafting | 13:56 |
jcastro | "oh wait, I like my job!" snip. | 13:56 |
rick_h | yea, it's kind of strange since the proejct/source is open and anyone can see it | 13:57 |
rick_h | will be nice when it's all past | 13:57 |
jcastro | security through "no one is deploying from trunk" | 13:57 |
rick_h | sinzui: ping, got a sec to chat? | 14:43 |
sinzui | rick_h, maybe in 5 | 14:45 |
* sinzui is with IS about canonistack environments | 14:45 | |
rick_h | sinzui: rgr, shoot me an invite when you're free please. I want to sanity check the upgrade path and such for adding this download field | 14:45 |
sinzui | rick_h, jcsackett: do either of you have time to review https://code.launchpad.net/~sinzui/charmworld/drain-ingest-queue/+merge/172353 | 15:44 |
rick_h | sinzui: I can in a sec | 15:45 |
rick_h | sinzui: trade ya https://code.launchpad.net/~rharding/charmworld/total-stats/+merge/172356 | 15:51 |
* sinzui looks | 15:51 | |
Makyo | jujugui call in 6 kanban now | 15:54 |
rick_h | sinzui: r=me | 15:54 |
Makyo | jujugui call in 2 - Gary's out; if anyone has a burning desire to run the call, cool. I will if not. | 15:58 |
sinzui | rick_h, I expect DISTRO_CHARM_DATA in charmworld/views/tests/test_charms.py to be updated. I think it documents which attributes are exposed by the old API. I believe we want to add the new attr and be certain we don't loose it when we drop old API. Am I wrong? | 16:03 |
rick_h | sinzui: looking | 16:03 |
rick_h | sinzui: I'm coming up empty with bzr grep DISTRO_CHARM_DATA | 16:04 |
rick_h | ? | 16:04 |
sinzui | oh, was it not landed... | 16:04 |
sinzui | ah! | 16:04 |
rick_h | I pulled trunk this morning? | 16:04 |
rick_h | or thought I did | 16:05 |
* rick_h goes to try agasin | 16:05 | |
sinzui | rick_h, not out problem. abentley cannot land his branch until exodus is available. He will need to update his branch since yours is r=me and can land now | 16:05 |
rick_h | sinzui: ah ok | 16:05 |
rick_h | jcastro: screenshot looks good in the blog post :) | 16:14 |
jcastro | YEAH! | 16:18 |
Makyo | Oh, forgot to mention in the call, did a blog post this morning about service placement. http://jujugui.wordpress.com/2013/07/01/focusing-on-services/ | 16:19 |
rick_h | sinzui: staging went boom it appears? | 17:32 |
* rick_h wonders if I did that :/ | 17:33 | |
* sinzui looks | 17:33 | |
sinzui | rick_h, it can go down for 15 minutes if it thinks it needs to do a full ingest and it was not previous configured to do so | 17:34 |
sinzui | has it been more than 15? | 17:34 |
rick_h | sinzui: not sure, just noticed it | 17:34 |
rick_h | sinzui: I figured there had been enough time post-merge for things to happen | 17:34 |
* sinzui visits the server because he has no oops in email | 17:34 | |
rick_h | sinzui: yea, it appears the wsgi app isn't running. apache is just dying | 17:35 |
sinzui | apache is running. I think you mean the the app is down | 17:36 |
sinzui | I see it stopped about 30 minutes ago | 17:36 |
rick_h | sinzui: right, wsgi (app, pyramid, etc) | 17:36 |
* sinzui tries to manually | 17:37 | |
sinzui | ingest is angry. | 17:37 |
* sinzui considers new dequeue power as a possible fix | 17:38 | |
rick_h | sinzui: :/ ok. I guess I did do something. /me tries locally again. | 17:38 |
sinzui | rick_h, I see this in the errors: http://pastebin.ubuntu.com/5817422/ | 17:39 |
rick_h | sinzui: yea, that's just a bad charm. It should carry on I thought. | 17:40 |
sinzui | I will dequeue, then restart to give ingest a chance to start | 17:40 |
rick_h | sinzui: rgr | 17:40 |
rick_h | sinzui: yea, it seems to run ok here locally so hoping I didn't break anything. | 17:41 |
sinzui | the webapp is up http://staging.jujucharms.com/ | 17:41 |
sinzui | search works | 17:41 |
rick_h | sinzui: cool, looks good. I just wanted to see the downloads stuff in the api ok | 17:41 |
sinzui | rick_h, I will look further at the logs to look for the root cause | 17:42 |
rick_h | sinzui: k, let me know if you think I need to peek at something. | 17:43 |
rick_h | jujugui jcsackett sinzui can I get a pair of reviews please? https://codereview.appspot.com/10832043 As noted, ignore the giant json files (2) and it's known that this won't QA against mjc at the moment. | 17:50 |
benji | rick_h: I'll lake a look. | 17:50 |
rick_h | benji: ty much | 17:51 |
rick_h | sinzui: and when you get a sec, want to verify what's next highest in priority please. | 17:51 |
sinzui | rick_h, there was an error in migration for the config-changed event, preventing the app from restarting | 17:52 |
sinzui | http://pastebin.ubuntu.com/5817453/ | 17:52 |
rick_h | sinzui: :/ ok. In my test it ran fine. Will follow up with a check for the other key first then. | 17:52 |
rick_h | sinzui: so that migration did not succeed. I'm safe to update it in place correct? | 17:53 |
sinzui | rick_h, I think we cannot assume the charm_data has the key. There are 5 charms that are invalid in the store, so cannot possibly have the value | 17:53 |
rick_h | sinzui: rgr | 17:53 |
rick_h | sinzui: yea, and in my limited set pulled down locally didn't have any of those | 17:53 |
sinzui | rick_h, You want to move your migration to the next number. Replace your old number with a no op. | 17:54 |
* sinzui looks at review | 17:54 | |
sinzui | rick_h, nm. | 17:55 |
rick_h | sinzui: on the no-op, nevermind? | 17:55 |
sinzui | rick_h, since your migration did not complete, I don't think migrations could have incremented the number. I think you can update your migration script as it is | 17:56 |
rick_h | sinzui: cool, that matches my thinking. | 17:56 |
rick_h | sinzui: did you want to sanity check this then? It's small enough that I'd like to get up vs reset db and pull down all of the charms in the db to try to get the bad ones | 18:00 |
rick_h | sinzui: https://code.launchpad.net/~rharding/charmworld/fix-downloads-migration/+merge/172386 | 18:00 |
sinzui | rick_h, Looks good | 18:14 |
rick_h | sinzui: cool | 18:15 |
sinzui | rick_h, do you want to see the staging oops? | 18:16 |
rick_h | sinzui: sure. I'm assuming it'll oops until the migraiton hits. | 18:16 |
rick_h | sinzui: guessing it's an attr not found issue on the sort from ES since the migration didn't pass muster | 18:16 |
sinzui | http://pastebin.ubuntu.com/5817523/ | 18:17 |
rick_h | sinzui: yea, that's the ES error if downloads isn't in the charm to do the order by for the /interesting url. Once this change lands and migrations take place should be good | 18:17 |
sinzui | ^ rick_h Since the server was offline for the migration, everything should be migrated. Ingest is not involved in the oops | 18:18 |
sinzui | of, I thought the change had just landed. | 18:18 |
sinzui | s/of/oh/ | 18:18 |
rick_h | sinzui: it's approved, but has to go through jenkins | 18:18 |
rick_h | sinzui: so should be a few minutes | 18:19 |
sinzui | oh, right, it does not claim to be merged | 18:19 |
Makyo | Remembered I totally had a dream about our comment style last night. I'm not sure how I feel about that. | 18:24 |
rick_h | Makyo: time to seek help | 18:24 |
sinzui | rick_h, LGTM | 18:24 |
rick_h | sinzui: thanks | 18:25 |
rick_h | here jenkins jenkins...here boy... | 18:29 |
rick_h | sinzui: branch is merged now it appears. Might get some error emails as I poke at staging again. | 18:57 |
* sinzui nods | 18:58 | |
rick_h | sinzui: ping, got a sec to help me out? Sorry, but can't get mutt to open the old ssh config email to access stagnig myself | 19:12 |
rick_h | sinzui: and I'd like to not leave things broken at EOD | 19:12 |
sinzui | staging fell over? | 19:13 |
rick_h | sinzui: well, the numbers aren't updating as I'd expect post-merge | 19:13 |
sinzui | oh | 19:13 |
* sinzui looks | 19:13 | |
rick_h | sinzui: and I can't get into staging to verify if it's still upgrading, the migration didn't run/errored again, etc | 19:13 |
* rick_h had hoped all would be well at this point | 19:14 | |
sinzui | oh, I think this is my fault | 19:14 |
sinzui | I did not clear the upgrade error from last effort | 19:14 |
sinzui | rick_h, sorry. this is my fault. I do need to address the charm error before tarmac can do its job | 19:16 |
rick_h | sinzui: ah, ok. Then I won't feel so bad then. Thanks for the heads up. | 19:17 |
rick_h | I've got to run and pick up the plate for the camper from the dealer before they close today, but I will check in tonight or tomorrow morning then. Thanks for the help today sinzui. Appreciate it. | 19:17 |
sinzui | rick_h, np. I will do the QA as best as I can given that I need your branch to kick the charm back to life | 19:19 |
benji | jujugui: I could use a couple of reviews for my drag and drop critical fix branch: https://codereview.appspot.com/10834043 | 19:34 |
sinzui | rick_h, everything looks good. After resolving, I used juju charmworld set revno=191 and waited for the reconfiguration: http://staging.jujucharms.com/api/2/charms/interesting shows me downloads and downloads_in_past_30_days. | 19:35 |
Makyo | benji, on it. Code LGTM, QAing. | 19:36 |
rick_h | sinzui: thanks! looks good now. | 20:57 |
sinzui | np | 20:57 |
rick_h | benji: around still? | 21:31 |
rick_h | benji: curious about your email and the two bits around config panel/"charm site being down" | 21:31 |
benji | rick_h: I got this message in a popup "Charm API server did not respond" | 21:31 |
rick_h | benji: ok, do you know what url that was? Was it consistent? We're not aware of any issues on the server today. Staging had issues during QA but nothing is pointing at it. | 21:33 |
benji | rick_h: XMLHttpRequest cannot load https://manage.jujucharms.com/api/1/charms/interesting. Origin http://localhost:8888 is not allowed by Access-Control-Allow-Origin. | 21:34 |
rick_h | benji: ah, please pull from trunk. api 1 has been deprecated about a week now and should be api2 | 21:34 |
rick_h | https://manage.jujucharms.com/api/2/charms/interesting should be the updated url and trunk should generate that for you. | 21:35 |
benji | rick_h: I was doing a binary search to find the point a regression was introduced, so that's not in the cards | 21:35 |
benji | I'll add this to my list of why having self-contained dev builds is good. | 21:35 |
benji | :/ | 21:36 |
rick_h | benji: ah, gotcha. :/ then we'll have to runa local charmworld instance. api1 has been removed from the server. So less that the server was down, and more that it's using an old version that won't be responding | 21:36 |
rick_h | benji: yea, I know jcsackett and I do that since we work on both parts of the code and often have to have them sync'd. | 21:36 |
rick_h | benji: please make sure to make some noise when you feel the server is 'down'. If it is and we've not said anything we need to jump on that and if it's something else we might be able to save you some time. | 21:37 |
benji | will do; thanks for the info | 21:37 |
rick_h | benji: so fyi, r272 removed api1 from the charmworld source. Running charmworld should be an easy bzr branch ... && make sysdeps && make install && make run | 21:41 |
rick_h | sinzui: we should probably deprecate old apis into a 'dead end' route that displays a warning as well. | 21:44 |
sinzui | +1 | 21:44 |
rick_h | sinzui: will file a bug real quick | 21:45 |
huwshimi | Morning | 22:59 |
rick_h | huwshimi: howdy | 23:38 |
rick_h | huwshimi: if you get a chance would like to chat on the controls stuff. | 23:39 |
huwshimi | rick_h: Sure, if you've got a minute | 23:46 |
rick_h | huwshimi: yea, let me get my headphones | 23:53 |
rick_h | huwshimi: invite on the way | 23:53 |
rick_h | darn it, how do you just create an empty hangout any more | 23:54 |
huwshimi | rick_h: Dunno, didn't get an invite | 23:54 |
rick_h | huwshimi: yea, you don't show up in my list so just trying to create an empty one but can't seem to figure out how to do that any more | 23:55 |
rick_h | huwshimi: https://plus.google.com/hangouts/_/cf491220dfb7736cee2876baf29110eeb52f5997?authuser=1 | 23:56 |
Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!