| *** matsuhashi has joined #openstack-trove | 00:05 | |
| *** jlucci has quit IRC | 00:20 | |
| *** jlucci has joined #openstack-trove | 00:20 | |
| *** jlucci has quit IRC | 00:36 | |
| *** krow has joined #openstack-trove | 00:40 | |
| *** asalkeld has left #openstack-trove | 00:41 | |
| *** yogeshmehra has joined #openstack-trove | 00:55 | |
| *** yogeshmehra has quit IRC | 01:02 | |
| *** nosnos has joined #openstack-trove | 01:05 | |
| *** erkules_ has joined #openstack-trove | 01:31 | |
| *** erkules has quit IRC | 01:34 | |
| *** robertmy_ has joined #openstack-trove | 01:40 | |
| *** robertmyers has quit IRC | 01:40 | |
| *** yogeshmehra has joined #openstack-trove | 01:59 | |
| *** yogeshmehra has quit IRC | 02:03 | |
| *** matsuhashi has quit IRC | 02:16 | |
| *** matsuhashi has joined #openstack-trove | 02:21 | |
| *** matsuhashi has quit IRC | 02:29 | |
| *** matsuhashi has joined #openstack-trove | 02:30 | |
| *** yogeshmehra has joined #openstack-trove | 03:00 | |
| *** dmakogon has quit IRC | 03:01 | |
| *** dmakogon has joined #openstack-trove | 03:01 | |
| *** yogeshmehra has quit IRC | 03:04 | |
| *** shivamS has joined #openstack-trove | 04:00 | |
| *** yogeshmehra has joined #openstack-trove | 04:01 | |
| *** yogeshmehra has quit IRC | 04:05 | |
| *** asalkeld has joined #openstack-trove | 04:32 | |
| *** asalkeld has left #openstack-trove | 04:32 | |
| *** vipul-away is now known as vipul | 04:38 | |
| *** zacksh_ has joined #openstack-trove | 04:42 | |
| *** shivamS has quit IRC | 04:47 | |
| *** zacksh has quit IRC | 04:49 | |
| *** shivamS has joined #openstack-trove | 04:49 | |
| *** shivamS has quit IRC | 04:50 | |
| *** shivamshukla has joined #openstack-trove | 04:59 | |
| *** SushilKM has joined #openstack-trove | 05:21 | |
| *** robertmyers has joined #openstack-trove | 05:43 | |
| *** robertmy_ has quit IRC | 05:43 | |
| *** yogeshmehra has joined #openstack-trove | 05:45 | |
| *** matsuhas_ has joined #openstack-trove | 05:57 | |
| *** matsuhashi has quit IRC | 05:58 | |
| *** nosnos_ has joined #openstack-trove | 05:58 | |
| *** nosnos has quit IRC | 05:59 | |
| *** robertmy_ has joined #openstack-trove | 06:30 | |
| *** robertmyers has quit IRC | 06:32 | |
| *** rushiagr has joined #openstack-trove | 06:35 | |
| *** matsuhas_ has quit IRC | 06:57 | |
| *** nosnos has joined #openstack-trove | 07:02 | |
| *** nosnos_ has quit IRC | 07:02 | |
| *** robertmyers has joined #openstack-trove | 07:05 | |
| *** robertmy_ has quit IRC | 07:05 | |
| *** matsuhashi has joined #openstack-trove | 07:05 | |
| *** yogeshmehra has quit IRC | 07:16 | |
| *** rushiagr has quit IRC | 07:36 | |
| *** matsuhashi has quit IRC | 07:40 | |
| *** rushiagr has joined #openstack-trove | 07:41 | |
| *** matsuhashi has joined #openstack-trove | 07:42 | |
| *** matsuhashi has quit IRC | 07:42 | |
| *** matsuhashi has joined #openstack-trove | 07:44 | |
| *** matsuhashi has quit IRC | 07:45 | |
| *** matsuhashi has joined #openstack-trove | 07:48 | |
| *** matsuhashi has quit IRC | 07:48 | |
| *** rushiagr has quit IRC | 08:07 | |
| *** matsuhashi has joined #openstack-trove | 08:13 | |
| *** yogeshmehra has joined #openstack-trove | 08:17 | |
| *** yogeshmehra has quit IRC | 08:21 | |
| *** robertmy_ has joined #openstack-trove | 08:27 | |
| *** robertmyers has quit IRC | 08:27 | |
| *** yogeshmehra has joined #openstack-trove | 08:27 | |
| *** yogeshmehra has quit IRC | 08:31 | |
| *** cweidenk1ller has quit IRC | 08:57 | |
| *** cweidenkeller has quit IRC | 08:58 | |
| *** ppenjoy has joined #openstack-trove | 09:13 | |
| *** yogeshmehra has joined #openstack-trove | 09:28 | |
| *** yogeshmehra has quit IRC | 09:32 | |
| *** ppenjoy has left #openstack-trove | 09:34 | |
| *** krow has quit IRC | 09:40 | |
| *** rushiagr has joined #openstack-trove | 10:10 | |
| *** SushilKM has quit IRC | 10:21 | |
| *** rushiagr_ has joined #openstack-trove | 10:26 | |
| *** rushiagr has quit IRC | 10:26 | |
| *** rushiagr_ is now known as rushiagr | 10:26 | |
| *** yogeshmehra has joined #openstack-trove | 10:29 | |
| *** ikhudoshyn has quit IRC | 10:32 | |
| *** yogeshmehra has quit IRC | 10:33 | |
| *** ikhudoshyn has joined #openstack-trove | 10:33 | |
| *** PradeepChandani has joined #openstack-trove | 10:57 | |
| openstackgerrit | Andrey Shestakov proposed a change to openstack/trove: Add support of UUID in flavors Removed int() conversion from flavors https://review.openstack.org/48735 | 11:09 |
|---|---|---|
| *** rushiagr has quit IRC | 11:11 | |
| *** asalkeld has joined #openstack-trove | 11:29 | |
| *** yogeshmehra has joined #openstack-trove | 11:30 | |
| *** yogeshmehra has quit IRC | 11:35 | |
| *** rushiagr has joined #openstack-trove | 11:46 | |
| *** SushilKM has joined #openstack-trove | 12:04 | |
| *** pdmars has joined #openstack-trove | 12:14 | |
| *** SushilKM has quit IRC | 12:16 | |
| *** SushilKM has joined #openstack-trove | 12:17 | |
| *** yogeshmehra has joined #openstack-trove | 12:30 | |
| *** SushilKM has quit IRC | 12:31 | |
| *** yogeshmehra has quit IRC | 12:35 | |
| *** ppenjoy has joined #openstack-trove | 12:42 | |
| *** robertmy_ has quit IRC | 12:44 | |
| *** ppenjoy has left #openstack-trove | 12:44 | |
| *** rushiagr has quit IRC | 12:49 | |
| *** jcru has joined #openstack-trove | 13:12 | |
| *** nosnos has quit IRC | 13:13 | |
| *** kevinconway has quit IRC | 13:13 | |
| *** rnirmal has joined #openstack-trove | 13:15 | |
| *** radez_g0n3 is now known as radez | 13:19 | |
| *** openstackgerrit has quit IRC | 13:22 | |
| *** openstackgerrit has joined #openstack-trove | 13:22 | |
| *** kevinconway has joined #openstack-trove | 13:32 | |
| *** haomaiwang has quit IRC | 13:32 | |
| openstackgerrit | Andrey Shestakov proposed a change to openstack/trove: Add support of UUID in flavors Removed int() conversion from flavors https://review.openstack.org/48735 | 13:33 |
| *** amytron has joined #openstack-trove | 13:43 | |
| *** Barker has joined #openstack-trove | 13:53 | |
| *** robertmyers has joined #openstack-trove | 14:03 | |
| *** demorris has joined #openstack-trove | 14:12 | |
| *** jlucci has joined #openstack-trove | 14:31 | |
| *** yogeshmehra has joined #openstack-trove | 14:32 | |
| *** yogeshmehra has quit IRC | 14:38 | |
| *** adrian_otto has joined #openstack-trove | 14:42 | |
| *** datsun180b has joined #openstack-trove | 14:46 | |
| *** jmontemayor has joined #openstack-trove | 14:53 | |
| *** adrian_otto has quit IRC | 14:56 | |
| *** adrian_otto has joined #openstack-trove | 14:57 | |
| *** jlucci1 has joined #openstack-trove | 15:04 | |
| *** Barker has quit IRC | 15:06 | |
| *** jlucci has quit IRC | 15:07 | |
| *** Barker has joined #openstack-trove | 15:11 | |
| openstackgerrit | Monty Taylor proposed a change to openstack/trove: Update from global requirements https://review.openstack.org/48986 | 15:16 |
| *** adrian_otto has quit IRC | 15:18 | |
| hub_cap | i will be merging this when robots +1 ^ ^ | 15:20 |
| *** paul_lodronio has joined #openstack-trove | 15:21 | |
| *** matsuhashi has quit IRC | 15:25 | |
| *** shivamshukla_ has joined #openstack-trove | 15:31 | |
| *** tanisdl has joined #openstack-trove | 15:33 | |
| ikhudoshyn | hub_cap: Hi, looks like there are busy days around? | 15:36 |
| hub_cap | HAH yes ikhudoshyn | 15:37 |
| isviridov | hub_cap, #link https://review.openstack.org/48986 does it mean that we can't use mockito anymore? | 15:37 |
| ikhudoshyn | how do you find new ios? (u seem to be the only one who cares<))) | 15:37 |
| ikhudoshyn | ,)) | 15:37 |
| hub_cap | ikhudoshyn: i hate it | 15:38 |
| ashestakov | hi | 15:38 |
| ashestakov | https://review.openstack.org/#/c/48735/ review pls | 15:38 |
| hub_cap | isviridov: im pretty sure mordred just removed ones we didnt use | 15:38 |
| ashestakov | hub_cap: any luck with service types? | 15:38 |
| ikhudoshyn | hub_cap: just like me, its kinda 'god plz let me unsee this') | 15:38 |
| hub_cap | not saying we "cant" use them | 15:38 |
| hub_cap | ikhudoshyn: yes i know | 15:38 |
| openstackgerrit | Monty Taylor proposed a change to openstack/trove: Update from global requirements https://review.openstack.org/48986 | 15:39 |
| mordred | oh. blech. I guess I should exclude test-requirements from that patch | 15:40 |
| ikhudoshyn | hub_cub: we now seem to be cool with Ed, i won't stay in his way. Now I'm doing some Mongo support. But before I'd like to merge some refactorings in GA. Just to extract some generic stuff from mysql specific code | 15:40 |
| ikhudoshyn | If u dont mind | 15:40 |
| mordred | but yes. you guys should have getting off of mockito and proboscis on your todo list | 15:40 |
| hub_cap | ashestakov: not yet its been a busy / frustrating wk | 15:40 |
| hub_cap | mordred: dont worry ;) | 15:40 |
| mordred | or otherwise convincing the rest of openstack to move to them | 15:40 |
| *** adrian_otto has joined #openstack-trove | 15:41 | |
| openstackgerrit | Monty Taylor proposed a change to openstack/trove: Update from global requirements https://review.openstack.org/48986 | 15:41 |
| hub_cap | mordred: /me is a standard whore and is working to make things like this standard... proby has been on my list, and now mockito is | 15:41 |
| mordred | there. left test-requirements out of that change, because I don't really care about that right now | 15:41 |
| ashestakov | hub_cap: maybe we can schedule time for review/commnts for service types? | 15:41 |
| hub_cap | ashestakov: im not toucing code today and ive got a full day | 15:42 |
| hub_cap | ive got a queue with your and dmakogon reviews in it | 15:42 |
| hub_cap | they will be done today. period. | 15:42 |
| mordred | hub_cap: I think you're just a whore in general :) | 15:42 |
| ashestakov | hub_cap: thanks | 15:43 |
| ikhudoshyn | hub_cup: plz add this https://review.openstack.org/#/c/48435/ to your queue. There is more to come, but I'd like to do it piece by piece | 15:43 |
| hub_cap | mordred: u know me too well | 15:43 |
| openstackgerrit | Monty Taylor proposed a change to openstack/trove: Update from global requirements https://review.openstack.org/48986 | 15:43 |
| mordred | one last time | 15:43 |
| hub_cap | ikhudoshyn: aye | 15:43 |
| mordred | maybe that will pass | 15:43 |
| ikhudoshyn | hub_cap: thanks | 15:44 |
| hub_cap | we shall see if you pass our unstandardness with your review mordred | 15:44 |
| * mordred punches hub_cap with a stick | 15:44 | |
| hub_cap | i mean, its like a gate of eliteness for trove. sure u can handle the rest of openstack, but have you seen THIS!!?!?!?! | 15:44 |
| * hub_cap vomits | 15:44 | |
| mordred | hub_cap: oh, you think you're the special ones? dude. SO MANY SNOWFLAKES | 15:45 |
| hub_cap | yup mordred dont put me on your tongue i might melt (wow that re-reads like a mordred comment) | 15:45 |
| datsun180b | i don't know why i'm dreading submitting a change to devstack so much | 15:48 |
| hub_cap | datsun180b: dont worry... be happy.... | 15:48 |
| * hub_cap is so cliché | 15:48 | |
| datsun180b | oh i'm pretty happy now. think i got the db access thing for conductor nailed this morning | 15:48 |
| hub_cap | datsun180b: its not that hard, the hard part is getting 2 people to +2 it | 15:48 |
| hub_cap | woot datsun180b | 15:49 |
| datsun180b | just have to make sure i can get an instance to go active without the logs screaming tracebacks at me | 15:49 |
| datsun180b | i've gotten the restart/retry process down to two scripts: ./IDDQD and then ./IDKFA | 15:49 |
| ikhudoshyn | datsun180b: hi, are there any news with conductor? | 15:50 |
| *** ashestakov has quit IRC | 15:50 | |
| datsun180b | ikhudoshyn: you just read it | 15:51 |
| hub_cap | lol datsun180b | 15:51 |
| ikhudoshyn | there's another guy who cant wait for it, i'll better go and teach him these spells | 15:51 |
| ikhudoshyn | ) | 15:52 |
| datsun180b | ikhudoshyn: if i can get an instance to go ACTIVE in this current test i'm running I'll submit my code reviews. That should be today | 15:52 |
| ikhudoshyn | oh! that will be great | 15:53 |
| ikhudoshyn | may the force be with u | 15:54 |
| *** demorris has quit IRC | 15:57 | |
| *** jmontemayor_ has joined #openstack-trove | 16:02 | |
| *** Nate2 has quit IRC | 16:04 | |
| *** jmontemayor has quit IRC | 16:04 | |
| *** tanisdl has quit IRC | 16:09 | |
| *** tanisdl has joined #openstack-trove | 16:10 | |
| *** Nate1 has joined #openstack-trove | 16:11 | |
| *** Nate2 has joined #openstack-trove | 16:14 | |
| *** Nate1 has quit IRC | 16:14 | |
| *** rnirmal has quit IRC | 16:15 | |
| cp16net | hub_cap: you mentioned you dont like the new ios? | 16:18 |
| cp16net | http://www.youtube.com/watch?v=JTpXVv-DaBQ | 16:24 |
| cp16net | new ios brings kids to tears | 16:24 |
| cp16net | lol | 16:24 |
| juice | oh-my-goodness | 16:26 |
| juice | man - the whiney, tired 5 year old was not in the user persona set at apple | 16:26 |
| * hub_cap watches cp16net | 16:26 | |
| hub_cap | wait | 16:26 |
| * hub_cap watches, cp16net | 16:26 | |
| cp16net | lol | 16:27 |
| hub_cap | tricky tricky commas | 16:28 |
| hub_cap | OMG thats the best 17 seconds ever | 16:28 |
| cp16net | hahaha | 16:29 |
| cp16net | :) | 16:29 |
| *** erkules_ is now known as erkules | 16:31 | |
| cp16net | juice: isnt that the truth... | 16:32 |
| *** ashestakov has joined #openstack-trove | 16:33 | |
| hub_cap | cp16net: juice this baby is way cooler http://www.youtube.com/watch?v=kOb1jesGECQ | 16:42 |
| cp16net | lol weird baby | 16:43 |
| hub_cap | ROUGH TOOTIN BABY | 16:44 |
| juice | i'm impressed that baby has such good balance on the hamper | 16:45 |
| juice | and the rhymes are superb | 16:46 |
| juice | assuming it was free-style | 16:46 |
| hub_cap | its gotta be freestyle | 16:48 |
| *** jmontemayor_ has quit IRC | 16:59 | |
| *** esp has joined #openstack-trove | 16:59 | |
| *** yidclare has joined #openstack-trove | 17:00 | |
| *** yogeshmehra has joined #openstack-trove | 17:03 | |
| *** Nate2 has quit IRC | 17:06 | |
| *** KennethWilke has joined #openstack-trove | 17:06 | |
| *** jmontemayor has joined #openstack-trove | 17:08 | |
| *** KennethWilke has quit IRC | 17:09 | |
| *** isviridov_ has joined #openstack-trove | 17:11 | |
| hub_cap | SlickNik: are the docs you started not on docs.o.o? | 17:13 |
| hub_cap | i dont see them (http://docs.openstack.org/) | 17:13 |
| esp | I think SlickNik is out today. | 17:19 |
| esp | hub_cap: could you please take a look at these at your convenience? https://review.openstack.org/#/c/48535/ https://review.openstack.org/#/c/47206/ | 17:24 |
| *** Nate1 has joined #openstack-trove | 17:26 | |
| *** isviridov_ has quit IRC | 17:26 | |
| hub_cap | esp: k | 17:26 |
| *** isviridov__ has joined #openstack-trove | 17:26 | |
| esp | hub_cap: thx! | 17:26 |
| *** adrian_otto has quit IRC | 17:26 | |
| *** jmontemayor has quit IRC | 17:28 | |
| *** demorris has joined #openstack-trove | 17:29 | |
| *** zigo_ has joined #openstack-trove | 17:32 | |
| zigo_ | Hi there! I'm trying to package Trove for Debian. I was wondering, what is trove-server for? Is this a central server, somehow??? | 17:33 |
| zigo_ | Come on guys, wake up! :) | 17:35 |
| *** isviridov_ has joined #openstack-trove | 17:35 | |
| juice | hello zigo_ | 17:35 |
| *** adrian_otto has joined #openstack-trove | 17:35 | |
| juice | where do you see a reference to trove-server? | 17:35 |
| zigo_ | juice: Hi. | 17:35 |
| zigo_ | juice: There's bin/trove-server in the source code. | 17:36 |
| juice | we have trove-api and trove-taskmanager and trove-guestagent | 17:36 |
| hub_cap | zigo_: hi | 17:36 |
| zigo_ | juice: Thanks a lot. | 17:36 |
| hub_cap | zigo_: trove-server is only for testing | 17:36 |
| hub_cap | just like juice says, the rest are either old artifacts or specifically for testing | 17:36 |
| *** isviridov__ has quit IRC | 17:37 | |
| zigo_ | For the moment, I have python-trove, trove-common, trove-api, trove-guestagent and trove-taskmanager as .deb files. Does this sound right? | 17:37 |
| hub_cap | trove-server is kinda like the other services, where you can spin up "all" services in one, like keystone / glance ( i dont know if they do that anymore, to tell u the truth...) | 17:37 |
| hub_cap | zigo_: thats how rackspace does it internally ;) | 17:37 |
| hub_cap | iirc | 17:38 |
| zigo_ | Oh, they already did some packaging work? Interesting. | 17:38 |
| hub_cap | well internally | 17:38 |
| zigo_ | It'd be nice if Rackspace was sharing a bit more ... :) | 17:38 |
| hub_cap | and likely wrong (i did some of it...) | 17:38 |
| zigo_ | I mean, on the packaging side. | 17:38 |
| datsun180b | we're trying | 17:38 |
| hub_cap | zigo_: ++ | 17:38 |
| zigo_ | datsun180b: Ah, well, then get in touch with me! :) | 17:38 |
| hub_cap | datsun180b: ya zigo_ is specific about packaging | 17:39 |
| zigo_ | datsun180b: I'm the guy maintaining all of OpenStack in Debian. | 17:39 |
| datsun180b | well i was speaking more generally | 17:39 |
| hub_cap | datsun180b: Thomas Goirand (look up his posts in openstack) | 17:39 |
| datsun180b | yeah /whois helps there | 17:40 |
| zigo_ | FYI guys, I'm currently working on Trove, then TrippleO and Marconi, since I'm done with the rest of (well, let's hope I am... :) ) | 17:40 |
| zigo_ | I guess I'll have more questions later on. | 17:40 |
| zigo_ | Anyway, thanks. | 17:40 |
| hub_cap | zigo_: great! ping us w/ questions | 17:42 |
| hub_cap | and ill see u in HK | 17:42 |
| hub_cap | not sure if u remember but we met in portland outside | 17:42 |
| zigo_ | hub_cap: I do remember I met some "reddwarf" people yes! :) | 17:42 |
| zigo_ | Not sure I would remember your face (met too many people). | 17:43 |
| hub_cap | i know zigo_, there is one of you, so i rmember heh | 17:43 |
| juice | zigo_ I might hit you up on packaging issues with debian vs pypi | 17:43 |
| zigo_ | hub_cap: I for sure will be happy to see you guys again. HK is like 2 hours away from my home ! :) | 17:43 |
| *** isviridov__ has joined #openstack-trove | 17:43 | |
| hub_cap | yup i knew you were in china i dont know where.. so ya ill see you in HK :) | 17:43 |
| juice | right now there is a nice unhealthy mix of the two types in openstack and it seems to cause recurring version issues | 17:43 |
| *** adrian_otto has quit IRC | 17:44 | |
| zigo_ | I live in Shanghai. | 17:44 |
| yogeshmehra | hubcap: i need few minutes on trove/heat | 17:45 |
| *** isviridov_ has quit IRC | 17:45 | |
| *** isviridov__ has quit IRC | 17:48 | |
| hub_cap | yogeshmehra: yes lets chat here | 17:48 |
| yogeshmehra | sure | 17:48 |
| yogeshmehra | was just checking.. | 17:49 |
| yogeshmehra | trove/heat, is it in progress... | 17:49 |
| yogeshmehra | i was not able to get the instances active... | 17:49 |
| *** jmontemayor has joined #openstack-trove | 17:49 | |
| hub_cap | yogeshmehra: no one is working on it exactly right now | 17:49 |
| yogeshmehra | and they are not reachable ssh | 17:49 |
| hub_cap | can you ping the IPs? | 17:49 |
| hub_cap | did you run the iptables MASQUERADE command? | 17:50 |
| hub_cap | just making sure of teh basics first | 17:50 |
| yogeshmehra | hub_cap: i did run the iptables MASQ... cmd | 17:50 |
| yogeshmehra | yeah, makes sense... | 17:50 |
| yogeshmehra | at one time, they were pingable...and nother, they were not | 17:50 |
| hub_cap | ok cool | 17:50 |
| hub_cap | did you add the heat key? | 17:51 |
| hub_cap | i forgot to add this in trove-integration | 17:51 |
| hub_cap | you might need to post a review for trove-integration | 17:51 |
| hub_cap | and add this | 17:51 |
| yogeshmehra | yeah, heatkey is done... | 17:51 |
| hub_cap | ok cool | 17:52 |
| zigo_ | OUTCH ! | 17:52 |
| hub_cap | try ssh ec2-user@ip | 17:52 |
| zigo_ | signing_dir = /tmp/keystone-signing-nova | 17:52 |
| hub_cap | zigo_: ya that makes sense right ;) ;) | 17:52 |
| zigo_ | This kind of joke generated a CVE for Nova! :) | 17:52 |
| hub_cap | can u file a bug for that zigo_? | 17:52 |
| zigo_ | Ok. | 17:53 |
| hub_cap | OUCH | 17:53 |
| yogeshmehra | hub_cap: it says cannot communicate on port 22 | 17:53 |
| hub_cap | not a good joke then!! | 17:53 |
| hub_cap | yogeshmehra: thats odd.. i have not tested in a while, can u do me a favor? | 17:53 |
| hub_cap | and you just try to create a instance using heat w/o trove on your setup | 17:53 |
| hub_cap | using our image | 17:53 |
| hub_cap | and see if it has the same result | 17:53 |
| yogeshmehra | hub_cap: creating an instance and even an instance group is success out of trove...directly using the heat api | 17:54 |
| *** isviridov_ has joined #openstack-trove | 17:54 | |
| yogeshmehra | image from within trove and using it to create an instancegroup works... | 17:55 |
| hub_cap | can you post the yaml you send to heat? | 17:55 |
| yogeshmehra | yeah... | 17:55 |
| hub_cap | oh this is nice... we should change our setup to use this | 17:55 |
| hub_cap | https://bugs.launchpad.net/heat/+bug/1101347 | 17:55 |
| yogeshmehra | it is: https://github.com/openstack/heat-templates/blob/master/cfn/F17/InstanceGroup.template | 17:55 |
| *** rnirmal has joined #openstack-trove | 17:55 | |
| yogeshmehra | this is the template i used for creating a 2 node deployement of an image from trove... | 17:56 |
| yogeshmehra | actually, the heat stack and the instance within the stack is visisble....but it doesn't come to ACTIVE state.. | 17:57 |
| hub_cap | but could you ssh in yogeshmehra? | 17:58 |
| hub_cap | thats the issue at hand | 17:58 |
| yogeshmehra | true...i couldn't | 17:58 |
| hub_cap | try to use a vanilla heat image | 17:59 |
| hub_cap | take our image and trove out of teh equation all together | 17:59 |
| hub_cap | make sure you have heat setup properly | 17:59 |
| yogeshmehra | hubcap: i have it setup... | 17:59 |
| yogeshmehra | i tested it using trove image itself...it did come up and was sshable... | 17:59 |
| yogeshmehra | Template: https://github.com/openstack/heat-templates/blob/master/cfn/F17/InstanceGroup.template and mysql image from trove... | 18:00 |
| hub_cap | lol @esp, you hit the RC1 oslo bug | 18:00 |
| esp | hmm | 18:00 |
| esp | do I win a prize? | 18:00 |
| esp | trip to hong kong? | 18:01 |
| hub_cap | you win a wait | 18:01 |
| esp | awesome. I win a lesson in patience. that's a more zen way to put it. | 18:02 |
| hub_cap | esp: look @ the comment i left | 18:02 |
| esp | k | 18:02 |
| *** jmontemayor has quit IRC | 18:04 | |
| esp | ah, ok. thx for clarifying! | 18:04 |
| *** isviridov_ has quit IRC | 18:05 | |
| *** demorris_ has joined #openstack-trove | 18:06 | |
| yogeshmehra | hub_cp: SOS.... :-) | 18:06 |
| SlickNik | hey gusy | 18:06 |
| SlickNik | guys* | 18:06 |
| yogeshmehra | i'll work out a fresh one and send you some logs.. | 18:06 |
| yogeshmehra | offline.. | 18:06 |
| SlickNik | hub_cap: travelling today, so will be on and off. | 18:07 |
| *** isviridov_ has joined #openstack-trove | 18:07 | |
| yogeshmehra | Slicknick: thanks.. | 18:07 |
| SlickNik | What's up yogeshmehra? | 18:07 |
| SlickNik | hub_cap: the docs that I put on are up on o.o | 18:07 |
| *** jmontemayor has joined #openstack-trove | 18:07 | |
| SlickNik | http://docs.openstack.org/ | 18:07 |
| *** demorris has quit IRC | 18:07 | |
| *** demorris_ is now known as demorris | 18:07 | |
| SlickNik | click on Python Developer Documentation | 18:08 |
| yogeshmehra | oops...i misread u... | 18:08 |
| hub_cap | ohhhhhhhh SlickNik thank you | 18:08 |
| hub_cap | i wsnt sure where | 18:08 |
| yogeshmehra | slicknick: forget thanks... :-) | 18:08 |
| *** SnowDust has joined #openstack-trove | 18:09 | |
| SnowDust | HeRe I CoMe ! | 18:09 |
| SnowDust | hello buddies .. | 18:10 |
| yogeshmehra | snowdust: hey | 18:10 |
| SnowDust | yogeshmehra: hi ! ! | 18:11 |
| yogeshmehra | i am figuring out the trove/heat... | 18:12 |
| yogeshmehra | will continue.. | 18:12 |
| SnowDust | ec2_user@ip wont work | 18:12 |
| SnowDust | no route to host is the answer :) | 18:13 |
| SnowDust | when u try ssh | 18:13 |
| SnowDust | that all means .. the port is blocked ? ? | 18:13 |
| yogeshmehra | is the instance pingable | 18:13 |
| SnowDust | nope | 18:13 |
| yogeshmehra | for u.. | 18:13 |
| SnowDust | only .. nova shows active status | 18:13 |
| SnowDust | with ip | 18:13 |
| yogeshmehra | it was for me once and not later.. | 18:13 |
| SnowDust | hmm .. | 18:13 |
| SnowDust | btw ... i read somewhere we have to make jeos images for heat ? | 18:15 |
| SnowDust | is that necessary ? | 18:15 |
| *** isviridov_ has quit IRC | 18:16 | |
| yogeshmehra | our vertica and mysql images work with heat | 18:16 |
| yogeshmehra | i booted them up using anotehr het template.. | 18:16 |
| yogeshmehra | but all this without trove.. | 18:16 |
| yogeshmehra | using heat apis | 18:16 |
| yogeshmehra | per vipul we should check the sec groups as well | 18:17 |
| yogeshmehra | for port 22 rule.. | 18:17 |
| SnowDust | tx ! | 18:18 |
| yogeshmehra | i am trying a new one.. | 18:19 |
| hub_cap | SnowDust: hello! | 18:19 |
| SnowDust | hi hub_cap | 18:19 |
| *** isviridov_ has joined #openstack-trove | 18:19 | |
| hub_cap | i see you asking questions all weekend, sorry that no one replied SnowDust. how are things coming? | 18:21 |
| *** isviridov_ has quit IRC | 18:21 | |
| SnowDust | i made my way upto .. the looping calls .. end of heat orchestration :) | 18:22 |
| SnowDust | then the box never comes to active state :) | 18:22 |
| SnowDust | and it timesout ( taskmanager ) | 18:22 |
| SnowDust | the instance is shown active in nova .. | 18:22 |
| SnowDust | heat says the stacking is complete | 18:22 |
| SnowDust | :) and with that me and yogeshmehra are trying to solve that .. | 18:23 |
| hub_cap | can you SSH into it SnowDust? | 18:24 |
| SnowDust | no .. as mentioned no route to host | 18:24 |
| SnowDust | i tried .. the vanilla installs .. no cocktail :d | 18:24 |
| SnowDust | still facing that .. | 18:25 |
| hub_cap | so even vanilla heat w/o any trove related stuff gives you no ssh ? | 18:26 |
| hub_cap | are you using the trove image? | 18:26 |
| hub_cap | try ssh ec2-user@IP too | 18:26 |
| yogeshmehra | hub_cap: without trove it does give ssh.... | 18:27 |
| hub_cap | ok | 18:27 |
| hub_cap | are you using cloud environments? like i could have a ssh into the machine to see whats going wrong/ or is it local to your laptops? | 18:28 |
| hub_cap | brb | 18:28 |
| SnowDust | i meant .. trove vanilla with heat configurations enabled .. in /etc/trove-taskmanager.conf and hitting | 18:29 |
| SnowDust | ./redstack rd-client instance create ... | 18:30 |
| SnowDust | that will run all the heat functions .. in /trove/trove/taskmanager/models.py | 18:30 |
| SnowDust | and .. will timeout .. during the poll activity to check machine status getting active | 18:30 |
| SnowDust | finally when i check .. | 18:30 |
| SnowDust | nova says box is active | 18:31 |
| SnowDust | heat says stacking is complete | 18:31 |
| SnowDust | and ssh says .. no route to host ! :-) | 18:31 |
| yogeshmehra | snowdust: per hub_cap, he wanted to know if without trove, the heat is functional | 18:31 |
| yogeshmehra | and it is... | 18:31 |
| SnowDust | yeah should be :) .. | 18:32 |
| SnowDust | or #heat will have meltdown ... | 18:32 |
| *** Barker has quit IRC | 18:33 | |
| *** jmontemayor has quit IRC | 18:34 | |
| *** isviridov_ has joined #openstack-trove | 18:35 | |
| *** isviridov_ has quit IRC | 18:37 | |
| *** Barker has joined #openstack-trove | 18:38 | |
| *** krow has joined #openstack-trove | 18:41 | |
| *** jlucci1 has quit IRC | 18:44 | |
| *** openstackgerrit has quit IRC | 18:47 | |
| *** openstackgerrit has joined #openstack-trove | 18:47 | |
| *** isviridov_ has joined #openstack-trove | 18:50 | |
| *** shivamshukla_ has quit IRC | 18:51 | |
| *** isviridov_ has quit IRC | 18:52 | |
| *** krow has quit IRC | 18:55 | |
| *** tanisdl has quit IRC | 18:55 | |
| redthrux | hey vipul, SlickNik - i wanted to get your take on https://review.openstack.org/#/c/48188/ | 18:55 |
| *** SnowDust has quit IRC | 18:56 | |
| *** paul_lodronio_ has joined #openstack-trove | 18:57 | |
| *** vipul has quit IRC | 18:57 | |
| *** vipul has joined #openstack-trove | 18:58 | |
| *** jlucci has joined #openstack-trove | 18:59 | |
| *** paul_lodronio has quit IRC | 18:59 | |
| *** paul_lodronio_ is now known as paul_lodronio | 18:59 | |
| *** jmontemayor has joined #openstack-trove | 19:00 | |
| *** krow has joined #openstack-trove | 19:03 | |
| *** radez is now known as radez_g0n3 | 19:05 | |
| *** tanisdl has joined #openstack-trove | 19:05 | |
| *** isviridov_ has joined #openstack-trove | 19:06 | |
| *** mmcdaris has joined #openstack-trove | 19:07 | |
| *** isviridov_ has quit IRC | 19:08 | |
| *** Nate1 has quit IRC | 19:09 | |
| hub_cap | ashestakov: bad news | 19:14 |
| hub_cap | after thinking a bit more about it | 19:14 |
| hub_cap | the change to flavors will cause a contract change | 19:14 |
| hub_cap | which means in order for it to become a string (its an int now) | 19:14 |
| hub_cap | we will have to rev the api | 19:14 |
| *** paul_lodronio has quit IRC | 19:17 | |
| *** paul_lodronio has joined #openstack-trove | 19:17 | |
| ashestakov | hub_cap: which contract? | 19:17 |
| *** isviridov_ has joined #openstack-trove | 19:22 | |
| *** isviridov_ has quit IRC | 19:25 | |
| demorris | ashestakov: the current API is running in production at RAX, so any non-additive changes to the V1.0 API will break the contract | 19:25 |
| demorris | it means that any changes that change it, have to be on V1.1 or V2.0 etc, thats all | 19:25 |
| *** KennethWilke has joined #openstack-trove | 19:26 | |
| demorris | it was inevitable that the project gets to this point, Nova has been through the same thing and it just means that Trove needs a strategy for API versioning, like all other openstack API's | 19:27 |
| ashestakov | hm, sounds bad when some contract restricts opensource | 19:27 |
| demorris | it's not restricting it at all | 19:27 |
| *** tanisdl has quit IRC | 19:27 | |
| demorris | changes can be made, you just can't change existing contracts | 19:27 |
| demorris | when you version an API, people then write bindings (python, C++, Java, etc.) and UI's, like horizon... | 19:28 |
| vipul | fwiw, I believe flavor "id" comes back as a string in the get/list instances, but comes back as an Int in flavor-list | 19:28 |
| demorris | if you change that, then you break those | 19:28 |
| datsun180b | it's more of a social contract: users already have expectation X from event Y | 19:28 |
| pdmars | and really, it has nothing to do with RAX, it's a contract between trove and its users | 19:28 |
| datsun180b | if Y is modified so that X is no longer fulfilled, we have broken the contract | 19:29 |
| demorris | correct, other providers, like HP could easily have this in production | 19:29 |
| datsun180b | if we implement new event Z, there was no previous expectation from it, and so we aren't necessarily breaking the contract | 19:29 |
| demorris | vipul: not sure where y'all are with that :) | 19:29 |
| vipul | demorris: :) | 19:30 |
| demorris | datsun180b: right, i consider those additive changes | 19:30 |
| datsun180b | i supposed if we offered a catalog API call then even additive changes could be considered breaky | 19:30 |
| demorris | we have seen the flavor id stuff change before (without catching it) and it breaks downstream users of Trove | 19:30 |
| datsun180b | but at that point i'm getting into hypotheticals | 19:30 |
| hub_cap | http://docs.rackspace.com/cdb/api/v1.0/cdb-devguide/content/POST_createInstance__version___accountId__instances_.html | 19:31 |
| hub_cap | "flavorRef": 1, | 19:31 |
| hub_cap | putting quotes around that breaks the contract | 19:31 |
| hub_cap | "flavorRef": "1" != "flavorRef": 1 | 19:31 |
| datsun180b | the flavor id type is a good example, we can't guarantee that someone wasn't relying that our flavors were ints in the json | 19:32 |
| demorris | yeah, and going back to what ashestakov, the point is not to be overly restrictive and limiting the project, these changes can be made, but the API needs to rev to do it, thats all | 19:32 |
| demorris | there are extremes to this too, the AWS approach (which is like the MS approach), is avoid revving the API ever | 19:34 |
| *** jmontemayor has quit IRC | 19:34 | |
| kevinconway | maybe instead of revving the API we should just rename the project when we make changes that violate the contract | 19:35 |
| kevinconway | that sounds easier | 19:35 |
| *** jmontemayor has joined #openstack-trove | 19:36 | |
| datsun180b | kevinconway: let's start by renaming you | 19:36 |
| hub_cap | kevinconway: LOL | 19:36 |
| *** tanisdl has joined #openstack-trove | 19:37 | |
| *** isviridov_ has joined #openstack-trove | 19:38 | |
| *** isviridov__ has joined #openstack-trove | 19:38 | |
| *** tanisdl_ has joined #openstack-trove | 19:40 | |
| *** isviridov__ has quit IRC | 19:40 | |
| *** tanisdl has quit IRC | 19:41 | |
| *** tanisdl_ is now known as tanisdl | 19:41 | |
| ashestakov | ok, how we can make changes in api? can we split next version api? | 19:41 |
| hub_cap | we make a 1.1 api ashestakov | 19:42 |
| hub_cap | evertying in 1.1 defaults back to 1.0 compatibility except for the changes | 19:43 |
| ashestakov | is it already done? | 19:43 |
| hub_cap | then we need 2x services, one in 1.1 and one in 1.0 | 19:43 |
| hub_cap | nope | 19:43 |
| hub_cap | its never been needed | 19:43 |
| hub_cap | the otehr services have it though (nova, glance...) | 19:43 |
| demorris | hub_cap: and those services dropped the point release concept | 19:43 |
| demorris | Nova went to just V1, V2, V3 naming convention | 19:43 |
| demorris | i believe | 19:43 |
| *** isviridov__ has joined #openstack-trove | 19:43 | |
| hub_cap | ashestakov: if u can make it so its an int or a string | 19:44 |
| hub_cap | then itll work | 19:44 |
| hub_cap | right demorris> | 19:45 |
| hub_cap | > | 19:45 |
| hub_cap | ? | 19:45 |
| hub_cap | HAH | 19:45 |
| *** Nate1 has joined #openstack-trove | 19:45 | |
| *** Nate1 has joined #openstack-trove | 19:46 | |
| demorris | I guess you could make the json accept both right? | 19:46 |
| demorris | i am not sure if that breaks any conventions | 19:47 |
| kevinconway | hub_cap: doesn't keystone serve two api revs from the same process? | 19:47 |
| hub_cap | yes thats how all the services work kevinconway | 19:47 |
| hub_cap | but its a question of do we want to do a ton of work for a smalll change if we can get _around_ it | 19:48 |
| ashestakov | can we do type conversion for flavors? | 19:49 |
| demorris | hub_cap: yeah, this change does not strike me as a sole reason to rev the API, so if there is a way to support it without breaking the contract, great | 19:49 |
| *** isviridov__ has quit IRC | 19:50 | |
| *** jlucci1 has joined #openstack-trove | 19:53 | |
| ashestakov | we can accept int values as well as strings, but do str(flavorid) for it | 19:53 |
| *** isviridov__ has joined #openstack-trove | 19:53 | |
| hub_cap | ashestakov: if we can accept ints and strings we should be able to tell the type from the json | 19:55 |
| *** isviridov__ has quit IRC | 19:55 | |
| hub_cap | the thing we have to avoid is this | 19:55 |
| hub_cap | currently you pass "flavorRef": 1 | 19:55 |
| hub_cap | if it becomes "flavorRef": "1" | 19:55 |
| hub_cap | its a contract change, going from an int to a string | 19:55 |
| hub_cap | even tho int("1") evals to 1 | 19:56 |
| *** jlucci has quit IRC | 19:56 | |
| ashestakov | i mean you still be able to pass "flavorRef": 1, but on server part it will convrt to str() | 19:56 |
| hub_cap | ashestakov: but json wont allow | 19:56 |
| hub_cap | "flavorRef": something_without_quotes_thats_not_an_int, | 19:57 |
| hub_cap | so | 19:57 |
| *** jlucci1 has quit IRC | 19:57 | |
| hub_cap | "flavorRef": 550e8400-e29b-41d4-a716-446655440000 | 19:57 |
| hub_cap | would fail in json | 19:57 |
| hub_cap | but | 19:57 |
| hub_cap | "flavorRef": "550e8400-e29b-41d4-a716-446655440000" would not | 19:57 |
| hub_cap | ok brb gotta watch my baby for a few minutes | 19:58 |
| ashestakov | is it covered by contract how you should pass values? with quotes or not | 19:58 |
| *** isviridov__ has joined #openstack-trove | 19:58 | |
| *** isviridov__ has quit IRC | 20:01 | |
| datsun180b | ashestakov: that would be in the schema of an individual request or response, so yes | 20:01 |
| datsun180b | one problem (in the case of flavor ids) is that xml doesn't so much have the types that json does, so whether flavor_id was a string or int was irrelevant so there was skew between our json and xml responses | 20:02 |
| *** isviridov__ has joined #openstack-trove | 20:05 | |
| *** isviridov__ has quit IRC | 20:07 | |
| hub_cap | if its possible, id allow both string and int | 20:10 |
| hub_cap | then we wont have to rev the api | 20:10 |
| *** isviridov__ has joined #openstack-trove | 20:10 | |
| hub_cap | is that possible w the schema code? | 20:10 |
| *** isviridov__ has quit IRC | 20:10 | |
| kevinconway | hub_cap: is it even possible to do both and not break the API contract? | 20:13 |
| kevinconway | because then it would be possible to get strings back out of the API | 20:13 |
| hub_cap | if its an int, itll return as such | 20:13 |
| hub_cap | if its a string, it will return as such | 20:13 |
| kevinconway | right, so your client has to know what got put in originally | 20:13 |
| hub_cap | im not sure that breaks the contract, due to the use case for each | 20:14 |
| hub_cap | i dont think thats the case | 20:14 |
| hub_cap | the server needs to know what it is in order to properly represent it | 20:14 |
| kevinconway | so if i load up my v1 client and it gets a UUID as the flavorRef, isn't that exactly what you're trying to prevent? | 20:14 |
| datsun180b | argh, i can't seem to get nova to play nice if i murder all screen sessions and try to install & kick-start again | 20:15 |
| hub_cap | kevinconway: if your installation uses UUID's then you would expect them | 20:15 |
| hub_cap | if it uses ints, you would expect those | 20:15 |
| hub_cap | if its using ints and strings... we could have issues | 20:15 |
| kevinconway | so you're saying it becomes a configuration option? | 20:15 |
| kevinconway | what i mean is that if I point my v1 client at any openstack deploy i would expect an integer back for flavorRef | 20:16 |
| hub_cap | https://gist.github.com/hub-cap/6769550 | 20:17 |
| hub_cap | remember we are talking creates here, so you would be submitting the flavorRef | 20:18 |
| *** jmontemayor has quit IRC | 20:18 | |
| hub_cap | and the list flavors will return json, and the id will be str/int | 20:18 |
| datsun180b | Explicit type checking in Python? That's a paddlin'. | 20:18 |
| hub_cap | and then you take that value and pass it in to the create | 20:18 |
| kevinconway | but if the id is ever str that would be an API break | 20:18 |
| hub_cap | datsun180b: care to write the compatibility layer for v1.1 ? | 20:18 |
| hub_cap | kevinconway: only for instances that require ints | 20:18 |
| datsun180b | Well it's not like you actually wrote if type(flavor_ref) == int | 20:19 |
| hub_cap | and the flavor list will return ints if thats the case | 20:19 |
| *** jmontemayor has joined #openstack-trove | 20:19 | |
| *** radez_g0n3 is now known as radez | 20:19 | |
| kevinconway | i guess i'm confused then. if we can have strings then why can't we have strings? | 20:19 |
| juice | just to be clear here - everything to nova is a string | 20:19 |
| juice | with respects to flavor id | 20:19 |
| juice | only in trove do we cast this is an int | 20:20 |
| juice | we store our reference (prior to ashketov_) change as an int | 20:20 |
| datsun180b | But since we made that distinction at one point in v1 we have to keep playing that game if we're going to claim we're still in v1. At least, that's what I understand from this. And in this case, the whole argument is moot when we accept and serve the request in XML | 20:21 |
| juice | so there is a mismatch currently between nova and trove | 20:21 |
| hub_cap | datsun180b: correct, but who does that anyway ;) | 20:21 |
| hub_cap | juice: correct | 20:21 |
| juice | datsun180b: this become tricky because we won't know whether they give us 1 or "1" | 20:21 |
| datsun180b | hub_cap: we have to leave the door open! what if they want an atom feed | 20:21 |
| juice | we could by convention do something like hub_cap proposed in his gist | 20:21 |
| datsun180b | or they send us GET http://legitimate/url/to/api/function.json | 20:22 |
| juice | or we can add a new field - flavor_id_description | 20:22 |
| juice | all strings go in description | 20:22 |
| juice | all ints continue to go into the current field | 20:22 |
| hub_cap | juice: naw its still the id field in nova | 20:23 |
| juice | over time we phase out the old int field | 20:23 |
| datsun180b | At one point we'll probably end up throwing the switch and flavors/<integer> will just 5XX | 20:23 |
| hub_cap | well we just need to rev the api to 1.1 and have backwards compat to 1.0 w/ a int, which is worst case | 20:23 |
| datsun180b | er, 4XX, user bad | 20:23 |
| hub_cap | datsun180b: yes thats the only problem | 20:24 |
| *** Nate1 has quit IRC | 20:24 | |
| hub_cap | is that the error returned will likely change | 20:24 |
| hub_cap | unless we add a flag (per kevinconway) to say "we only allow ints" | 20:24 |
| juice | i like my plan the best :) | 20:24 |
| hub_cap | which id rather rev teh api at that point | 20:24 |
| cp16net | wow.... | 20:24 |
| datsun180b | my last job had a problem when they wanted to migrate to a new platform and shake up the api in a big way | 20:24 |
| cp16net | http://youtu.be/Q0W0MR6Ybgk?t=26s | 20:24 |
| hub_cap | i dont think its hella work to reve the api | 20:24 |
| datsun180b | after lots of customer education and such eventually they just built a monster and paraded the new api wearing the skin of the old api | 20:25 |
| hub_cap | ok so, executive decision | 20:25 |
| hub_cap | rev the api | 20:25 |
| hub_cap | its too hacky to maintain compat | 20:25 |
| datsun180b | eventually they swapped out, i understand, but there was an interim | 20:26 |
| hub_cap | if we rev the api, then rax (and hp) can decide when they go to v1.1 | 20:27 |
| hub_cap | or tehcnically probably v2 :) | 20:27 |
| hub_cap | ill talk to some people who have rev'd the api's in nova | 20:27 |
| datsun180b | this is kind of weird, but why do the screens in trove all start in /opt/stack/python-troveclient | 20:27 |
| datsun180b | we must be missing a popd | 20:27 |
| kevinconway | is this the only change we want for 1.1 or do we want to batch them for the rev? | 20:27 |
| hub_cap | kevinconway: the changes just stay in v1.1 | 20:27 |
| hub_cap | if we have a 2nd change, we put it in v1.1 w/ backwards compat | 20:28 |
| hub_cap | the rest of v1.1 is a passthru to v1.0 (acutally thats written backwards) | 20:28 |
| vipul | what happens to say /instances.. is that only in v1.0 or visible in 1.1 | 20:28 |
| hub_cap | everything is in 1.1, and all changes to the api | 20:28 |
| vipul | oh i guess you anaswered that | 20:28 |
| hub_cap | let me splain | 20:28 |
| hub_cap | well kinda | 20:28 |
| *** Nate1 has joined #openstack-trove | 20:28 | |
| hub_cap | so everything is promoted to 1.1, and any calls to 1.0 are routed to 1.1 | 20:28 |
| hub_cap | unless they have a explicit 1.0 service | 20:29 |
| hub_cap | thats how i think itll work | 20:29 |
| hub_cap | then the 1.0 can just quote the int and push it (likely it wont even need to, python right!!) | 20:29 |
| hub_cap | push it to the 1.1 rev | 20:29 |
| vipul | hmm.. but everything is in 1.0 today.. so you'd move the current stuff to v1.1? | 20:29 |
| hub_cap | itd be in both 1.0 and 1.1 | 20:29 |
| hub_cap | the v flag is in the paste.ini only | 20:30 |
| vipul | oh cuz you said "unless they have a explicit 1.0 service" | 20:30 |
| hub_cap | so it really doesnt matter... | 20:30 |
| hub_cap | correct | 20:30 |
| vipul | seems like everything woudl have a 1.0 service | 20:30 |
| hub_cap | well the common api_factory can define that i think | 20:31 |
| hub_cap | so basically we need to figure it out :) | 20:31 |
| hub_cap | let me ask some people | 20:31 |
| vipul | yea i get the concept I think.. so not really a full Rev of the API | 20:31 |
| vipul | just an extension.. dot release | 20:32 |
| hub_cap | i think so | 20:32 |
| hub_cap | mabye ;) | 20:32 |
| hub_cap | cuz the other projects just use /v1 /v2 | 20:32 |
| hub_cap | and default to v1.1 in the background | 20:32 |
| vipul | ah ok | 20:32 |
| hub_cap | so ll have to figure out how to go back to 1.0 | 20:32 |
| hub_cap | *we'll have to figure out | 20:32 |
| hub_cap | this will require a pretty massive rewrite id assume, of the api handling code | 20:33 |
| * hub_cap wonders if its tiem to move to pecan/wsme in the process | 20:33 | |
| vipul | everyone's doing it :D | 20:34 |
| hub_cap | it _is_ the future of OpenStack | 20:36 |
| * hub_cap puts it on the todo list | 20:36 | |
| hub_cap | problem is... adding flavors in horizon adds them as uuid's and ashestakov couldnt use them | 20:36 |
| vipul | ouch | 20:37 |
| vipul | we could do something crazy, like detect if it's an int or string in API code | 20:38 |
| hub_cap | LOL vipul | 20:38 |
| hub_cap | scroll up | 20:38 |
| vipul | oh | 20:38 |
| vipul | :P | 20:38 |
| hub_cap | https://gist.github.com/hub-cap/6769550 | 20:38 |
| hub_cap | after chatting w/ kevinconway and co, it just became obviously hacky | 20:39 |
| vipul | well then.. that should go in v1.0 | 20:39 |
| datsun180b | Is it worth it to make a distinction between what would go in a minor versus major version change | 20:42 |
| hub_cap | it depends on how the other openstack projects do it | 20:43 |
| hub_cap | they dont seem to offer > 1 minor version | 20:43 |
| datsun180b | like requiring minor changes might just build on the previous version a la class APIv1_7(APIv1_6) but not hold that requirement for major versions | 20:43 |
| datsun180b | but if the answer is lockstep then forget it | 20:43 |
| datsun180b | i think i'm getting caught up in irrelevant details while trying to help with this 1.0/1.1 situation | 20:44 |
| hub_cap | the answer is "the standards gods of OpenStack (mordred)" want us to keep in line.. and frankly i dont blame them | 20:45 |
| hub_cap | the family that plays together stays together! | 20:45 |
| datsun180b | oh of course, so my musings are irrelevant as speculation | 20:45 |
| lifeless | hub_cap: openstack is basically semver | 20:45 |
| lifeless | hub_cap: but openstack has a nasty tendancy to backwards incompat changes in api servers every release | 20:46 |
| hub_cap | i dont want to do that lifeless | 20:46 |
| lifeless | hub_cap: as features are dropped after deprecation periods | 20:46 |
| datsun180b | i'm not about to try to give you cat herding pointers | 20:46 |
| hub_cap | and as much as i see they semver, the apis are only single major based :( | 20:46 |
| lifeless | sure | 20:46 |
| hub_cap | so i guess they have just dropped their compat for the _last_ minor by now | 20:47 |
| hub_cap | if they ever had it ;) | 20:47 |
| hub_cap | and datsun180b fwiw, this is a 1.1 vs 1.0 thing | 20:48 |
| *** ashestakov has quit IRC | 20:49 | |
| hub_cap | we would have backward compat between 1.1 and 1.0 for this 'feature' | 20:49 |
| datsun180b | yeah i'm trying to help but i'm all thumbs here | 20:50 |
| hub_cap | this will help for a general understanding http://semver.org/ | 20:51 |
| *** radez is now known as radez_g0n3 | 20:53 | |
| datsun180b | were you serious about that backward compatibility question earlier | 20:53 |
| hub_cap | hmm what question? | 20:54 |
| datsun180b | <hub_cap> member:datsun180b: care to write the compatibility layer for v1.1 ? | 20:54 |
| datsun180b | fyi colloquy is terrible for irc | 20:54 |
| hub_cap | ahhhh datsun180b no i wasnt | 20:54 |
| hub_cap | focus on the conductor | 20:54 |
| datsun180b | oh i'm focusing | 20:55 |
| mordred | hub_cap: what was the question? | 20:55 |
| datsun180b | if i could get this business to stop crying when i kill screen and restart it would speed things up, but right now i'm minutes from making sure i've got the db updated properly and then it's frantic code cleanup time | 20:56 |
| mordred | oh - API version | 20:56 |
| hub_cap | mordred: "Have you ever drunk Baileys from a shoe?" | 20:56 |
| hub_cap | no, boo on api version! | 20:56 |
| mordred | lifeless: I do not believe we are semver or even close to it for API versioning. | 20:56 |
| mordred | I think we may be getting close to thinking about wanting to be there for library stuff | 20:56 |
| *** mordred has quit IRC | 20:57 | |
| *** mordred has joined #openstack-trove | 20:57 | |
| * datsun180b suppresses urge to suggest something ouroborean as a joke | 20:58 | |
| hub_cap | :) | 20:58 |
| hub_cap | mordred: we (Trove) can still use it as a guideline | 21:00 |
| hub_cap | i dont want to provide non compat to minor versions (but it seems like we, OpenStack, punt on that and just major rev everything) | 21:01 |
| lifeless | mordred: The intent I've seen on all reviews for apis is 'don't break it', 'new version if it would break it' | 21:01 |
| lifeless | mordred: which isn't that far off of semver, though I take your point | 21:01 |
| mordred | hub_cap: I think if you come up with a GOOD way to not screw people with your API versioning, no-one will be unhappy | 21:02 |
| *** pdmars has quit IRC | 21:03 | |
| hub_cap | heh yes mordred | 21:03 |
| *** jcru has quit IRC | 21:12 | |
| *** yogeshmehra has quit IRC | 21:31 | |
| *** isviridov_ has quit IRC | 21:32 | |
| *** adrian_otto has joined #openstack-trove | 21:33 | |
| *** robertmyers has quit IRC | 21:41 | |
| *** yogeshmehra has joined #openstack-trove | 21:45 | |
| *** Barker has quit IRC | 21:48 | |
| *** yogeshmehra has quit IRC | 21:58 | |
| hub_cap | vipul: https://review.openstack.org/#/c/48986/ <-- lets merge this | 22:06 |
| *** amytron has quit IRC | 22:10 | |
| *** amytron has joined #openstack-trove | 22:10 | |
| *** datsun180b has quit IRC | 22:11 | |
| *** datsun180b has joined #openstack-trove | 22:11 | |
| *** paul_lodronio has quit IRC | 22:11 | |
| *** krow has quit IRC | 22:13 | |
| *** amytron has quit IRC | 22:17 | |
| *** rnirmal has quit IRC | 22:23 | |
| *** vipul is now known as vipul-away | 22:26 | |
| *** krow has joined #openstack-trove | 22:31 | |
| *** vipul-away is now known as vipul | 22:35 | |
| *** KennethWilke has quit IRC | 22:36 | |
| *** demorris has quit IRC | 22:38 | |
| *** jmontemayor has quit IRC | 22:45 | |
| *** shakayumi has joined #openstack-trove | 23:00 | |
| *** mmcdaris has quit IRC | 23:03 | |
| hub_cap | thx for the approve vipul | 23:04 |
| vipul | np hub_cap | 23:04 |
| vipul | hub_cap: Ok with me merging https://review.openstack.org/#/c/47206/? | 23:07 |
| hub_cap | most def vipul | 23:07 |
| vipul | <3 | 23:08 |
| datsun180b | "status": "ACTIVE", | 23:09 |
| datsun180b | IT IS DONE | 23:09 |
| vipul | weeeee | 23:09 |
| datsun180b | it's taken so long to get just that to work and i think i may have found a bug in servicestatus.from_description in the process | 23:09 |
| hub_cap | datsun180b: can u imagine trove saying i am the alpha and the omega | 23:09 |
| esp | thx! | 23:09 |
| datsun180b | now to fix this bidness once and for all | 23:10 |
| hub_cap | horray | 23:10 |
| datsun180b | you'll hear from jenkins when i submit those reviews | 23:13 |
| *** datsun180b has quit IRC | 23:13 | |
| openstackgerrit | A change was merged to openstack/trove: PEP8. F841 https://review.openstack.org/46289 | 23:18 |
| *** mmcdaris has joined #openstack-trove | 23:18 | |
| openstackgerrit | A change was merged to openstack/trove: Update from global requirements https://review.openstack.org/48986 | 23:22 |
| openstackgerrit | A change was merged to openstack/trove: Added server_id to my.cnf template logic https://review.openstack.org/47206 | 23:22 |
| *** tanisdl_ has joined #openstack-trove | 23:31 | |
| *** tanisdl has quit IRC | 23:33 | |
| *** tanisdl_ is now known as tanisdl | 23:33 | |
| *** asalkeld has quit IRC | 23:42 | |
| *** yogeshmehra has joined #openstack-trove | 23:43 | |
| hub_cap | vipul: SlickNik can someone help me w/ this? | 23:45 |
| hub_cap | volume_time_out is SUPER low in the config (cfg.py) | 23:45 |
| hub_cap | and when you try to prov an instance using the test env, it fails miserably | 23:46 |
| hub_cap | but why is it not failing the gate? | 23:46 |
| hub_cap | redthrux: around? | 23:46 |
| hub_cap | cfg.IntOpt('volume_time_out', default=2), | 23:47 |
| hub_cap | thats 2 seconds LULZ | 23:47 |
| vipul | must be that fast | 23:50 |
| vipul | there is a sleep of 2.. which may make it 4 seconds | 23:50 |
| openstackgerrit | Michael Basnight proposed a change to openstack/trove: Volume timeout is set too low https://review.openstack.org/49089 | 23:51 |
| vipul | volume_time_out=30 | 23:51 |
| vipul | that's what's defined in sample | 23:51 |
| hub_cap | oh i guess thats what int tests use | 23:53 |
| hub_cap | the problem is when we dont use int-tests | 23:53 |
| hub_cap | we have a broke-ass build | 23:53 |
| hub_cap | but this is good | 23:53 |
| hub_cap | cuz moving to devstack is finding issues we have in our config | 23:53 |
| hub_cap | https://review.openstack.org/#/c/49089/ | 23:53 |
| hub_cap | i bet server_delete_time_out is too low too | 23:53 |
| vipul | i assume in devstack we just take our .sample and use that? | 23:53 |
| hub_cap | nope | 23:54 |
| hub_cap | dtroyer insisted we do not do that | 23:54 |
| hub_cap | so we can find issues like this in our code | 23:54 |
| vipul | so in devstack there is no sample? | 23:54 |
| hub_cap | nope | 23:54 |
| hub_cap | its inisets | 23:54 |
| hub_cap | to a blank file | 23:54 |
| hub_cap | w/ devstack you get like ~6 values | 23:54 |
| vipul | Oh | 23:54 |
| hub_cap | in the tmgr conf | 23:54 |
| vipul | that explains that | 23:55 |
| hub_cap | yup :) | 23:55 |
| hub_cap | ill up the server_delete_time_out=480 too | 23:55 |
| hub_cap | but to 60 | 23:55 |
| hub_cap | 480 is crazy big | 23:55 |
| vipul | that could be what yogeshmehra was seeing with Heat.. | 23:55 |
| openstackgerrit | A change was merged to openstack/trove: Fixed method signature _create_server_volume_heat https://review.openstack.org/48535 | 23:55 |
| vipul | although that would not be subject to this timeout | 23:55 |
| openstackgerrit | Michael Basnight proposed a change to openstack/trove: Volume timeout is set too low https://review.openstack.org/49089 | 23:55 |
| hub_cap | i wonder if those are stupid low cuz of tests | 23:56 |
| hub_cap | if so ill cry | 23:56 |
| hub_cap | LITERALLY CRY | 23:56 |
| *** adrian_otto has quit IRC | 23:56 | |
| hub_cap | "lets tune down our defaults for test config" = RIDICULOUS FAIL | 23:56 |
| *** adrian_otto has joined #openstack-trove | 23:56 | |
| vipul | well i know we've turned things UP for tests.. not sure if we've turned things down | 23:57 |
| hub_cap | fake tests | 23:57 |
| vipul | oh yea those | 23:57 |
| hub_cap | to make them faster | 23:57 |
| hub_cap | ya | 23:57 |
| hub_cap | so if the fake tests come back at like +1min slower | 23:58 |
| hub_cap | we will know thats the case, and i will cry | 23:58 |
| *** matsuhashi has joined #openstack-trove | 23:58 | |
| vipul | I wonder.. does Heat also implement such polling timeouts? | 23:58 |
| hub_cap | im sure it has polling timeouts.. it'd have to | 23:59 |
| hub_cap | maybe things dont stay stuck in build anymore | 23:59 |
| hub_cap | if thats the case it doesnt need to | 23:59 |
| vipul | well for things like volumes.. seems like you want Heat to do the same thing | 23:59 |
| vipul | wait for it to be 'available' before booting an instance | 23:59 |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!