[04:37] morning all === stokachu_ is now known as stokachu === vladk|offline is now known as vladk [06:45] jam: morning [06:46] morning vladk. note that you probably don't have to work today, since you should get 2 swap days for travelling. (at least, I don't think you've taken both days yet) [06:46] jam: yes, I am still recovering today [06:47] vladk: how was your flight back? (uneventful I hope) [06:49] jam: all was fine [06:50] vladk: good to hear. Are you suffering from jetlag at all? [06:51] jam: unfortunately yes, I terribly want to sleep now [06:52] vladk: ouch. I know for me the best way to switch is on the flight itself. For some reason I can generally force myself to rest at the appropriate sleeping time of the destination, and that really helps reset my clock. [06:52] if that doesn't work, then make sure you get sunshine when you can, as I've heard that helps, too. [06:54] jam: sunshine is good, but it's rain now in Moscow and only +4 === vladk is now known as vladk|offline [10:46] morning natefinch are you coming to the standup? [10:46] I was hoping fwereade would show up, but he isn't on IRC [10:46] though he is responding to emails === bodie_ is now known as Guest50187 === bodie_ is now known as Guest46720 [15:15] Ey! I see a new line in juju status: state-server-member-status: has-vote [15:16] what is this magic?! [15:16] did we get quorem?! [15:20] lazyPower: that's HA (high availability) [15:21] ahh ok. i thought that was crazy fast to land a quorem feature. [15:21] lazyPower: if you're on a non-local provider, you can do juju ensure-availability -n 3 and it'll bring up two more state servers === tvansteenburgh1 is now known as tvansteenburgh [15:21] ahhh ok. was any mention of that posted to the list or changelog? [15:21] lazyPower: you mean like leader election, or something else? Not sure what quorem means wrt juju [15:22] natefinch: leader election. there's a few charms in teh store that would benefit from that. [15:22] lazyPower: it's on the roadmap, but yeah, not implemented yet. Most of juju devs are still recovering from the sprint ;) [15:23] natefinch: +1 from me for recovering [15:23] sorry for the excitement :P I saw that and was instantly energized to figure out what it's wrt [15:24] lazyPower: as for announcing HA... I had thought there was an announcement, but possibly not. That was one of the things we talked about during the sprint - better communication of what features we're working on and what we're releasing [15:25] natefinch, hey, are you the only one around at this time today? seems everyone is taking holidays [15:25] right on. You guys are doing a bang up job. Thanks for rocking it [15:25] dimitern: o/! [15:25] hey lazyPower ;) [15:26] dimitern: yeah, pretty much just me. Everyone else gets two swap days. [15:26] natefinch, i'll post a review for you to see then shortly [15:26] dimitern: cool. How come you're not taking a second swap day? [15:27] natefinch, i think with some help from the server guys containers on ec2 can get addresses and talk to other instances and v.v. [15:27] natefinch, i'm in austin still [15:27] natefinch, but i'm taking next week of [15:27] off [15:27] dimitern: ahh, didn't realize you were still in the US. [15:28] natefinch, yep, i've got my share of the us for this year i hope :) [15:28] dimitern: a week in Vegas is enough of the US for anyone :) [15:30] natefinch, and denver, and now austin :) === marcoceppi_ is now known as marcoceppi [16:18] bac, Orangesquad credentials and swift are happy again. I am still investigating bug 1312217. Make sure all envs have control-bucket and access-secret specified in environments.yaml to avoid breaking swift access [16:18] <_mup_> Bug #1312217: open-stack provider breaks swift with standard config === Guest46720 is now known as bodie_ [16:57] is there a way to get a unique ID for a mongo document through the transaction mechanism? [16:57] for a new document* [16:57] or do you just use the built in labix/mgo stuff [16:57] I just don't want to step on any toes [17:09] am I asking the wrong question? is there somewhere obvious I'm not looking? [17:10] bodie_: do you mean inserting a new document in a transaction and getting back a mongo generated id? [17:16] right [17:16] bodie_: the only way I see it maybe possible is using the txn.Runner ChangeLog [17:17] bodie_: still looking [17:17] bodie_: http://docs.mongodb.org/manual/tutorial/create-an-auto-incrementing-field/ [17:17] http://docs.mongodb.org/manual/tutorial/create-an-auto-incrementing-field/ [17:17] jcw4: too slow ;) [17:17] right... but everything in the State document is dealt with through a special handler called txn [17:18] bodie_: http://docs.mongodb.org/manual/reference/glossary/#term-objectid [17:18] damnit no... I copied your link and accidentally pasted it in irc [17:18] jcw4: heh [17:18] ugh... yes... that's as in "do you just use the built in labix/mgo stuff" [17:18] natefinch: :) [17:18] bodie_: looks like the right way to do it is to let mongo set your id. Which works unless you need to know the id before you commit [17:19] so that would be using Upsert? or is it implicit? [17:19] precisely, and since the ID is built into the document, I need to compose the transaction using that ID. [17:19] natefinch: and if that is in a transaction then to retrieve that _id requires the changelog? [17:19] or we need some other mechanism for generating a uuid for the Action documents [17:20] I suppose one way would be to generate the document using a transaction, then look it up and upsert it with the ID that was generated for it. but that seems so kludgy [17:20] bodie_: you mean use the generated id in another document within the same transaction? [17:21] bodie_: http://godoc.org/labix.org/v2/mgo/bson#NewObjectId [17:21] bodie_: that's what you want to use [17:22] I was hoping to use this http://paste.ubuntu.com/7405686/ [17:22] * bodie_ checks what natefinch said... was looking up how I'm doing it now [17:22] okay, so we can bypass the transaction mechanism to use newObjectID? [17:23] bodie_: they're not really related. The object Id is just a unique id. You can set whatever you want as the document's _id as long as it's unique [17:23] right, that's what I'm getting at lol [17:23] I didn't see any other technique for coming up with a unique identifier except I guess a hash function or something [17:23] bodie_: yes, you can use NewObjectId for the id and still use the documents in a transaction [17:24] alright, thanks for the assist :) [17:26] good news sinzui [18:10] natefinch, bug 1316709 could be marked low (we don't think we will support linux mint series) or wont fix (we don't want to support linux mint in clouds) or invalid (The user could set default-series in the yaml) [18:10] natefinch, your thoughts? [18:17] sinzui: likely the intent is just to use the client on mint, which seems reasonable. I think saying that they should set default-series is a good workaround. I'd say won't fix with the default-series workaround is ok [18:18] thank you natefinch [18:52] natefinch, quick & easy review? https://codereview.appspot.com/91150044/ [18:54] dimitern: sure [18:54] natefinch, ta! [19:17] jcw4, I'm going to implement that test, so hopefully basics on Action will be good to go by EOD [19:17] bodie_: cool [19:24] natefinch, thanks for the review [19:24] dimitern: welcome [21:11] o/ menn0 [21:12] thumper: howdy :) [22:04] * thumper goes to make a coffee so he can trawl email with coffee in hand [22:09] thumper, menn0 hey ;) [22:17] dimitern: hi! [22:51] down to under 300 unread work emails [22:57] thumper: heh, I only am missing the review ones