19:01:15 #startmeeting Poppy Weekly Meeting 19:01:16 Meeting started Thu Jul 30 19:01:15 2015 UTC and is due to finish in 60 minutes. The chair is amitgandhinz. Information about MeetBot at http://wiki.debian.org/MeetBot. 19:01:17 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 19:01:19 The meeting name has been set to 'poppy_weekly_meeting' 19:02:01 #topic RollCall 19:02:21 o/ 19:03:06 o/ 19:03:08 o/ 19:03:11 o/ 19:03:16 o/ 19:03:41 #link Agenda: https://wiki.openstack.org/wiki/Meetings/Poppy 19:03:45 #topic Recap 19:03:52 #link http://eavesdrop.openstack.org/meetings/poppy_weekly_meeting/2015/poppy_weekly_meeting.2015-07-23-19.02.html 19:04:03 amitgandhinz to make mimic fastly the default driver 19:04:34 this is still a WIP - i have the mimic driver code working in poppy. but i found many endpoints were not implemented in the fastly mimic api 19:04:44 so many of the tests failed against it 19:04:57 i should be able to get it passing by EOD tomorrow 19:05:01 woot, the mimic driver works! 19:05:18 #action amitgandhinz to make mimic fastly the default driver 19:05:22 it will help us in starting to run api tests at the gate. 19:05:32 yup, thats the goal 19:05:46 i also want to add the dns mimic driver that does not exist yet 19:05:57 else we can run the tests at the gate against the Default DNS driver 19:06:10 malini to get the outstanding security patches fixed and merged 19:06:25 still WIP - we had one patch merged this week 19:06:31 will continue working with HEnry 19:06:41 #action malini to get the outstanding security patches fixed and merged 19:06:51 malini to clean up/fix the many test related patches that have been out there for a while 19:06:58 this is going to be an ongoing task - security tests 19:07:04 We should make this more specific 19:07:15 still messy there :/ 19:07:31 #malini to deal with https://review.openstack.org/143720 19:07:33 oops 19:07:38 #action malini to deal with https://review.openstack.org/143720 19:07:47 thanks - lets do one a week :) 19:07:48 #action malini to deal with https://review.openstack.org/144324 19:07:53 NOOOOOOO 19:07:56 lets start with these two 19:07:58 haha 19:07:58 haha 19:08:01 ok - not bad 19:08:05 yelling helps 19:08:08 either fix or abandon 19:08:30 I can get these tow done in a week :) 19:08:45 cool 19:09:01 #action amitgandhinz to assign more things to malini at next meeting 19:09:05 grrr.... 19:09:09 hehe 19:09:15 tonytan4ever to hound people 4 ever to review his stuff 19:09:22 I have been doing that 19:09:34 and he got his ip merged 19:09:35 didnt we merge the IP restriction? 19:09:35 yaya 19:09:40 yep 19:09:43 To get the IP restrictions merged, but still need to do that get notification driver merged 19:09:44 I have to add an E2E test 19:09:45 hounding worked!! 19:09:54 well hounded 19:10:16 I would say keep this item on as I am rolling out more patches. 19:10:26 #topic Liberty 2 19:10:30 #link https://launchpad.net/poppy/+milestone/liberty-2 19:10:32 guess what ! 19:10:39 its release time! 19:10:45 liberty 3 starts tomorrow 19:10:47 Woot!!! 19:10:57 i'll try to cut a release today after this meeting 19:11:23 these cycles feel like they are too short 19:11:42 tonytan4ever: https://blueprints.launchpad.net/poppy/+spec/ip-restrictions 19:11:45 this is merged right 19:11:48 Right 19:12:05 tonytan4ever: https://blueprints.launchpad.net/poppy/+spec/geo-restrictions 19:12:23 This one I'd say put it in review 19:12:25 wow, time runs by fast. 19:12:37 already is 19:13:00 Cool 19:13:00 tonytan4ever: https://blueprints.launchpad.net/poppy/+spec/akamai-ssl-driver 19:13:07 Started? 19:13:07 This one is in progress. 19:13:25 or Good Progress 19:13:31 i'll do good progress 19:13:36 Yes I started it. Good progress. 19:13:56 A notification driver bp is coming along with it. 19:14:23 tonytan4ever: https://blueprints.launchpad.net/poppy/+spec/notification-driver 19:14:31 ok let me link them 19:14:35 Woot, it is in review. 19:15:43 https://bugs.launchpad.net/bugs/1433807 19:15:44 Launchpad bug 1433807 in Poppy "Reenable shared ssl + add domains/origins PATCH tests" [Medium,New] 19:15:44 Launchpad bug 1433807 in poppy "Reenable shared ssl + add domains/origins PATCH tests" [Medium,New] 19:15:49 anyone know the status of this bug? 19:16:01 Launchpad bug 1433807 in poppy "Reenable shared ssl + add domains/origins PATCH tests" [Medium,New] https://launchpad.net/bugs/1433807 19:16:26 * amitgandhinz is there an echo in here? 19:16:29 I think I have dealt with it a while back 19:16:44 It should not be an issue anymore. 19:16:44 good question - I might have added tht when we disabled it for some reason 19:16:59 we can try running it, and see if it fails. 19:17:01 tonytan4ever: do we have those api tests? 19:17:01 malini: can you check, and if so we can close the bug 19:17:08 ok..checking now 19:17:55 https://bugs.launchpad.net/poppy/+bug/1421183 19:17:56 Launchpad bug 1421183 in Poppy "PATCH - Adding a second rule to an origin causes the service to go to failed state" [High,New] 19:17:56 Launchpad bug 1421183 in poppy "PATCH - Adding a second rule to an origin causes the service to go to failed state" [High,New] 19:17:58 Launchpad bug 1421183 in poppy "PATCH - Adding a second rule to an origin causes the service to go to failed state" [High,New] https://launchpad.net/bugs/1421183 19:18:02 cool. so fyi liberty-3 will start tomorrow, and go on until Sep 3rd 19:18:04 we need to look at that as well. 19:18:30 looks like we have the shared ssl test https://github.com/stackforge/poppy/blob/master/tests/api/services/data_patch_service.json#L307 19:18:33 sriram: i will add that to lib-3 19:18:45 but idk if it still exists 19:18:49 feels like it shouldnt 19:18:54 amitgandhinz: I think we fixed that at some point. 19:19:02 I can check and report back on that. 19:19:20 malini: line 307-317 in data_patch_service.json should cover that case 19:19:41 sriram: I dont see a corresponding patch test for the second rule 19:19:46 Can you add tht? 19:19:53 malini: sure 19:19:57 malini: thanks i added the link to the comment and will close the bug 19:20:09 I just checked locally, and the bug doesnt seem to exist any more. 19:20:53 ok updated them both 19:21:25 https://bugs.launchpad.net/poppy/+bug/1420945 19:21:27 Launchpad bug 1420945 in Poppy "default ttl not automatically assigned" [Medium,Confirmed] - Assigned to Amit Gandhi (amit-gandhi) 19:21:28 Launchpad bug 1420945 in poppy "default ttl not automatically assigned" [Medium,Confirmed] https://launchpad.net/bugs/1420945 19:21:32 Launchpad bug 1420945 in poppy "default ttl not automatically assigned" [Medium,Confirmed] 19:21:39 sriram: this is done also right? 19:21:50 amitgandhinz: yep 19:22:11 I didnt address the bug in the commit message … :'( 19:22:33 & nobody caught that :-$ 19:22:38 baah. 19:23:18 anyway we can assign that to me, and mark as done. 19:23:28 this is also done: https://bugs.launchpad.net/poppy/+bug/1466908 19:23:29 Launchpad bug 1466908 in Poppy "Patch Updates don't appear immediately" [Undecided,New] - Assigned to Obulapathi (obulpathi) 19:23:46 Launchpad bug 1466908 in poppy "Patch Updates don't appear immediately" [Undecided,New] 19:23:47 Launchpad bug 1466908 in poppy "Patch Updates don't appear immediately" [Undecided,New] https://launchpad.net/bugs/1466908 19:24:48 how bug list is getting shorter =) 19:24:52 i like =D 19:25:22 I think this one was fixed 19:26:05 this one didnt address the bug either.. 19:26:11 https://bugs.launchpad.net/poppy/+bug/1443619 19:26:12 Launchpad bug 1443619 in Poppy "restrictions rule with no request_url returns 500" [Medium,Incomplete] 19:26:12 we really need to watch for that. 19:26:14 amitgandhinz: din't see you message up 19:26:22 Launchpad bug 1443619 in poppy "restrictions rule with no request_url returns 500" [Medium,Incomplete] 19:26:23 Launchpad bug 1443619 in poppy "restrictions rule with no request_url returns 500" [Medium,Incomplete] https://launchpad.net/bugs/1443619 19:26:37 never seen the uvirtbot before. 19:27:28 sriram: Its the siri of IRC 19:28:00 is that 500 still an issue? 19:28:36 there should be some validation in there right? 19:28:44 anyway this is easy to investigate. 19:28:56 I can take a look a this as well and report back 19:29:33 the patch the bug refers to is abandoned 19:29:59 marking it as invalid unless sriram repro's it 19:30:15 roger that 19:30:45 obulpathi: can you explain this bug - https://bugs.launchpad.net/poppy/+bug/1434609 19:30:46 Launchpad bug 1434609 in Poppy "PoppyCDN pulls content from origin server by default" [Undecided,Incomplete] 19:30:46 Launchpad bug 1434609 in poppy "PoppyCDN pulls content from origin server by default" [Undecided,Incomplete] 19:30:48 Launchpad bug 1434609 in poppy "PoppyCDN pulls content from origin server by default" [Undecided,Incomplete] https://launchpad.net/bugs/1434609 19:31:11 amitgandhinz: IIRC you fixed the 500 bug in another patch 19:31:16 isnt the point of cdn to pull content from an oriign? 19:31:37 If a user creates a service with no ttl rules 19:31:59 the content is pulled from the origin server all the times 19:32:08 obulpathi: if the content is not cacheable, then it always goes to origin (unless it has a caching ttl) 19:32:29 and now we are also adding a default ttl to services (which the user can remove if they want) 19:32:51 delete it? 19:32:56 i will mark it as invalid 19:32:57 yep 19:33:00 this should be addressed by the default ttl 19:33:17 yep, the default ttl also takes care of this 19:33:30 amitgandhinz: please mark it as invalid 19:33:49 https://bugs.launchpad.net/poppy/+bug/1443979 19:33:50 Launchpad bug 1443979 in Poppy "API Tests - Negative tests fail for the wrong reason" [Medium,New] 19:33:55 malini: this was fixed right? 19:33:57 nope 19:34:05 Launchpad bug 1443979 in poppy "API Tests - Negative tests fail for the wrong reason" [Medium,New] https://launchpad.net/bugs/1443979 19:34:06 Launchpad bug 1443979 in poppy "API Tests - Negative tests fail for the wrong reason" [Medium,New] 19:34:06 we validate for 400s on the negative tests 19:34:28 But we had a lot of code changes & now the tests get 400 due to different reasons 19:34:34 not what the test says it is 19:34:42 ok 19:35:26 we rename the test then 19:35:29 job over lol 19:35:37 ideally we shud validate the resp message for 400s 19:36:01 +1 ensure that the 400 is for what we expect the 400 to be and not a different reason 19:36:13 yes, +1 19:36:18 https://bugs.launchpad.net/poppy/+bug/1473193 19:36:20 Launchpad bug 1473193 in Poppy "Prevent use of root domains in services" [Undecided,Confirmed] - Assigned to Obulapathi (obulpathi) 19:36:20 but tht is a lot of work & I am not sure where it fits in the priorities 19:36:21 Launchpad bug 1473193 in poppy "Prevent use of root domains in services" [Undecided,Confirmed] 19:36:30 obulpathi: you were working on this right? 19:36:34 obulpathi had a patch for this. 19:36:39 Launchpad bug 1473193 in poppy "Prevent use of root domains in services" [Undecided,Confirmed] https://launchpad.net/bugs/1473193 19:36:48 Yep, I had put it on shelve while working on Kafka 19:37:01 I will unshelve it and finish the failing API tests 19:37:13 the PR is out tehre tough 19:37:16 cool - i will assign it to lib3 19:37:25 thanks :) 19:37:50 obulpathi: I am sure you'll finish it much earlier ;) 19:37:50 https://bugs.launchpad.net/poppy/+bug/1408301 19:37:51 Launchpad bug 1408301 in Poppy "PATCH Requests for Domains Returns Confusing Error Message" [Medium,New] 19:37:58 Launchpad bug 1408301 in poppy "PATCH Requests for Domains Returns Confusing Error Message" [Medium,New] 19:37:59 Launchpad bug 1408301 in poppy "PATCH Requests for Domains Returns Confusing Error Message" [Medium,New] https://launchpad.net/bugs/1408301 19:38:08 I think I have encountered this as well. 19:38:30 the error message from jsonpath was confusing? 19:38:35 its basically a bad json patch 19:38:35 jsonpatch 19:38:45 all of our jsonpatch messages are confusing :/ 19:38:54 but should we be handling it a bit more gracefully? 19:39:05 we need a schema to evaluate schemas!! 19:39:06 :P 19:39:11 and we sometimes have verbal diarrhea in the error responses 19:39:12 :D 19:39:30 they do look ugly 19:39:56 thats because of the embedded traceback right? 19:40:05 or do you mean something else? 19:40:16 amitgandhinz: ^ 19:40:28 this is because jsonpatch validation error messages are ugly 19:40:47 it makes sense for jsonpatch to return those messages 19:40:52 sometimes its the traceback. sometimes it the jsonpatch messages dumping a crapload of json to say a key is wrong 19:40:55 But we probably should use them better 19:41:07 like it will dump the whole json schema to say you can only enter one item of something 19:41:22 It's kind tough to translate those messages to be a prettier form though./ 19:41:26 it would be nicer to customize those messages but i dont think you can with jsonpatch 19:41:31 tonytan4ever: +1 19:41:51 amitgandhinz: looks like thats a separate project all together 19:41:52 we should convert all those messages to 'ooops -try again later' 19:41:58 ok im done with bugs, any one else want to bug the team about bugs? 19:42:15 some kind of egress middleware decorator 19:42:16 * amitgandhinz for some reason i thought malini was singing britney spears 19:42:24 :D 19:42:25 amitgandhinz: what? 19:42:31 lol 19:42:41 hahaha 19:43:02 amitgandhinz: anyway if we did that, it will be a separate project, we plug into poppy. 19:43:02 #topic Open Discussion 19:43:03 if we extend out current validations in out code, we can have better error messages 19:43:18 did everyone vote yet? 19:43:23 who hasnt voted? 19:43:27 obulpathi: correct, but I think the scope is far broader for this. 19:43:27 me 19:43:31 go vote! 19:43:34 sriram: true 19:43:37 here are the links 19:43:38 does stoplight have the functionality? 19:43:39 I voted! 19:43:44 done voting. 19:43:55 amitgandhinz: ^ 19:44:11 my voting done! 19:44:16 • CDN for your Cloud - Openstack Poppy https://www.openstack.org/summit/tokyo-2015/vote-for-speakers/Presentation/3882 19:44:16 • Getting High with Poppy CDN https://www.openstack.org/summit/tokyo-2015/vote-for-speakers/Presentation/5170 19:44:16 • Money Laundering with Poppy CDN https://www.openstack.org/summit/tokyo-2015/vote-for-speakers/Presentation/5175 19:44:18 • 19:44:19 might need to look into it, long time since I looked at stoplight. 19:44:20 Avoiding the 7 Deadly Sins while engineering your applications on Openstack https://www.openstack.org/summit/tokyo-2015/vote-for-speakers/Presentation/5506 19:44:26 malini: i dont have your ones handy 19:44:41 amitgandhinz: I havent voted for anybody's :/ 19:44:51 GO VOTE! 19:44:55 GO VOTE! 19:44:55 vote vote vote!! 19:44:57 GO VOTE!GO VOTE!GO VOTE!GO VOTE!GO VOTE!GO VOTE!GO VOTE!GO VOTE!GO VOTE! 19:44:59 GO VOTE!GO VOTE!GO VOTE!GO VOTE!GO VOTE!GO VOTE!GO VOTE!GO VOTE!GO VOTE! 19:45:00 I just voted. 19:45:00 I don't feel so bad since it includes mine :d 19:45:01 GO VOTE!GO VOTE!GO VOTE!GO VOTE!GO VOTE!GO VOTE!GO VOTE!GO VOTE!GO VOTE! 19:45:19 amit213: lol 19:45:23 damn it 19:45:31 amitgandhinz: heh 19:45:42 GOVOTE is that Italian? 19:45:46 https://www.openstack.org/summit/tokyo-2015/vote-for-speakers/presentation/4463 19:45:49 * sriram must master tab completion 19:46:03 https://www.openstack.org/summit/tokyo-2015/vote-for-speakers/presentation/4470 19:46:29 I think I voted - but I dont see the count go up 19:46:39 where do you see counts? 19:46:56 those are dummy counts - nvm :/ 19:47:02 found mine https://www.openstack.org/summit/tokyo-2015/vote-for-speakers/presentation/4470 19:47:09 i linked it above lol 19:47:20 4463 and 4470 19:47:27 :D 19:47:38 voted. 19:48:13 voting ends today 19:48:16 so go vote! 19:48:43 #topic Highwinds CDN 19:48:51 if you did an inverted index on this chat history, i'm sure vote is at No.1 19:48:57 :P 19:49:05 so on an unrelated topic, had a great discussion with highwinds yesterday 19:49:09 intellectual! 19:49:12 they are interested in contributing to Poppy 19:49:18 any questions (lol) 19:49:20 cymbals!! 19:49:23 that is good to know 19:49:24 yay! 19:49:31 will be building the driver? 19:49:32 #topic Akamai CDN 19:49:35 timelines? 19:49:41 #topic Akamai and Highwinds 19:49:48 s/will/will they 19:49:49 ok lets cover them both 19:49:55 ok 19:50:11 so highwinds is interested, they have the resources and will hopefully start working with us to build their own driver 19:50:19 yayyy 19:50:21 i think they are stil ramping up 19:50:34 so hopefully they engage and participate 19:50:43 lets make sure we welcome them to our channel 19:50:50 of course 19:50:56 Akamai is also interested in taking over the akamai driver 19:51:02 woot..woot 19:51:09 tonytan4ever: ^^ 19:51:15 i had a 1.5 hour discussion purely focused on Poppy with them today 19:51:20 so we just sit back & watch everybody work <:o) 19:51:23 they seemed very keen and really like the idea of it 19:51:25 That sounds good. 19:51:44 they had a few questions around features that others didnt offer and how to integrate that into poppy 19:52:02 but they were happy with my answers around only exposing common features via the api 19:52:14 and anything else are either backend defaults, or we dont offer it 19:52:31 poppy exposes only the most common set of features, over different providers right. 19:52:37 we may need to look into flavor settings to enable certain features that some providers have but not all 19:52:45 amitgandhinz: +1 19:52:55 ie, lets say akamai, cloudfront, and cloudflare all offer ddos protection 19:53:03 but fastly, maxcdn, and highwinds dont 19:53:24 we could offer a feature in the flavor settings that enables certain features only (or make them as extensions?) 19:53:34 it makes sense. 19:53:36 yeah makes sense 19:53:50 so we still prevent vendor lock in or vendor specific features, but its no longer the lowest common denominator 19:53:52 But wouldn't that require us to store some feature info as well ? 19:54:01 maybe 19:54:03 thats fine 19:54:08 as long as its in a flavor 19:54:09 it almost becomes like feature flags 19:54:17 +1 for feature flags 19:54:23 K, just probably another field of flavor table 19:54:28 possibly 19:54:28 yeah we can switch flavors whenever needed 19:54:31 we need to think about it 19:54:36 so dont really see a lock in there. 19:54:49 as long as the api remains compatible we are okay 19:54:55 yeah 19:55:05 what we dont want is endpointA is for providerA and endpointB is only for providerB 19:55:27 hmmmm 19:55:32 we kinda have tht today :/ 19:55:38 where? 19:55:43 malini: example? 19:55:55 * amitgandhinz 4 min warning 19:56:00 that probably needs to re-thought of then. 19:56:05 IRC didnt like tht 19:56:11 admin/provider/akamai/service 19:56:16 we can move to openstack-poppy if we exceed timelimit 19:56:20 ah admin specific endpoints 19:56:27 that doesnt really count 19:56:32 those are for operators only 19:56:33 booo 19:56:43 users would never hit those endpoints 19:57:00 fair 19:57:12 yeah those will probably be shielded from users 19:58:21 ok, great meeting 19:58:21 ok anything else to discuss? 19:58:25 thats all folks 19:58:29 see ya 19:58:31 #endmeeting