21:09:11 #startmeeting Solum team meeting 21:09:11 Meeting started Tue Dec 16 21:09:11 2014 UTC and is due to finish in 60 minutes. The chair is devkulkarni. Information about MeetBot at http://wiki.debian.org/MeetBot. 21:09:12 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 21:09:15 The meeting name has been set to 'solum_team_meeting' 21:09:22 #topic Roll call 21:09:30 james li 21:09:32 devdatta kulkarni 21:09:36 gil pilz 21:09:38 Akshay Chhajed 21:09:39 ravi sankar penta 21:09:43 murali allada 21:09:45 hi ravips 21:09:46 long time 21:09:52 hi dev 21:09:54 hi akshay 21:09:56 hi gil 21:10:07 hi dev 21:10:16 hello devkulkarni 21:10:28 lets wait for a minute to see if anyone else joins us 21:10:34 Melissa Kam 21:10:41 hey mkam 21:11:26 #topic announcements 21:11:42 does anyone have any announcements? 21:11:56 ravips: hope this time works for you 21:12:06 yes 21:12:12 cool 21:12:16 how about you akshay? 21:12:46 i am stuck with different error now 21:12:59 I meant does the time work for you? 21:13:11 we will come to the error part soon 21:13:11 * ravips taking time off next week and won't be able to attend the meeting 21:13:18 ya.... i am fine with this time 21:13:32 okay.. cool.. thanks for the heads up ravips 21:13:57 okay.. lets move on to the action items 21:14:04 #topic Review Action items 21:14:12 #link http://eavesdrop.openstack.org/meetings/solum_team_meeting/2014/solum_team_meeting.2014-12-09-21.08.txt 21:14:44 james_li, muralia: can we try to push this forward https://review.openstack.org/#/c/140162/ 21:14:53 this was one of the action items from last week 21:14:58 sure 21:15:06 thanks muralia 21:15:51 were there any other action items that folks were carrying forward? 21:16:10 I remember I had one about identifying low hanging fruit bugs 21:16:16 I did that sometime back 21:16:41 muralia: what is the latest on the juno release? i remember you were coordinating that with adrian 21:17:07 He's still working on it. I'm waiting for him to reply with a confirmation email. 21:17:16 okay, cool 21:17:18 I know that he has been pon vacation for a few days 21:17:36 devkulkarni: I think that patch is ready to go 21:17:56 james_li: cool.. will take a look 21:18:22 okay, if there are no other action items I will go to next topic 21:18:29 #topic Blueprint/Task review 21:18:37 In this I have one 21:18:49 it is ravi's patches 21:18:59 ravips: did you see the comments on your patches? 21:19:18 yes, I did 21:19:48 do they make sense to you? any thoughts? 21:20:47 after we have done discussing ravi's patches, I would like to consider errors that akshayc is running into 21:20:48 yes, definitely..I will respond to those comments 21:21:23 I'm bit busy on my other tasks and won't be able to get to those soon...I would like to handover this task to someone who is interested in this task..any volunteers? 21:21:42 ravips: I see 21:22:17 if we don't get volunteers immediately that should be alright imo 21:22:44 ravips: how about this? 21:23:14 once you have responded to the comments, we can outline a path of what needs to happen next in the reviews 21:23:31 sounds like a plan 21:23:32 and then mark it has an open wishlist bug for new contributors 21:23:50 cool 21:23:50 yep.... this will be good.... 21:24:18 are there any disagreements? 21:24:26 +1 21:24:29 +1 21:25:31 #agreed ravips will respond to comments on his patches. devkulkarni, ravips, and others will outline in details what needs to happen next as part of the review. we will seek volunteers from the team to complete the work. 21:25:57 next up is the issue that akshayc is running into 21:26:06 akshayc: floor is yours 21:26:49 ok. i receive a 502 after ==> devstack: Status: Image is up to date for solum/slugtester:latest 21:27:07 i checked the code..... i can see the UI page.... 21:27:28 but it gets stuck when i provide the github link 21:27:29 do you have detailed stack trace? what screen this is showing up in? worker? 21:28:06 error is vagrant up devstack command 21:28:26 oh.. you mean your vagrant up itself is failing? 21:28:40 yep. 21:28:55 but the demo UI is visible but not functional 21:29:10 but where are using specifying the github link? 21:29:12 where does the demo UI log? 21:29:16 oh you are trying the demo UI? 21:29:25 yep.... 21:29:29 it used to be functional till Atlanta summit. 21:29:41 I have the same problem 21:29:43 after that we haven't really gone back and test it 21:29:51 gpliz: you are also trying the demo UI? 21:30:01 s/test/tested/ 21:30:06 the vagrant devstack does seem to be useable despite these failures 21:30:27 vagrant devstack is definitely very useful 21:30:27 dev - no, I'm using the dev env devstack 21:30:40 akshayc: gpilz: what errors are we seeing in vagrant provision/build logs (under /opt/stack/logs/) 21:30:52 gpilz: okay. so you are seeing the same problem as akshayc in your devstack setup? 21:30:56 yes 21:31:02 okay 21:31:21 ravips: they are not even getting there 21:31:34 what I'm trying to say is that, even if you get those errors 21:31:49 you should try to use the env anyway 21:31:54 right 21:32:00 yes 21:32:07 akshayc: can I suggest something.. 21:32:31 ya.... i was trying the same... despite the errors.... i see the UI.... but it gets stuck 21:32:32 once you do vagrant up, even if it fails, you should be able to login to the VM using vagrant ssh 21:33:00 once inside the VM, try source the openrc (source ~/devstack/openrc) and then ./stack.sh 21:33:11 s/source/sourcing/ 21:33:25 devkulkarni: ./unstack.sh and then ./stack.sh will be clean 21:33:25 I would not spend time with the demo UI 21:33:28 2014-12-16 20:52:58.655 | stack.sh completed in 1668 seconds. 21:33:37 ravips: sure 21:33:41 ok 21:33:48 akshayc: that is great if stack.sh completed 21:34:11 can you login to a screen session and see if all the solum services are running correctly or not? 21:34:22 screen -x stack should drop you into a screen session 21:35:16 are nova, keystone, glance, swift, neutron, heat, solum* processes up and running? 21:35:19 there should be 36 things in screen 21:35:27 yep.... 21:35:28 right 21:35:36 okay 21:35:52 i see .... .* 32$(L) solum-api 33$(L) solum-build-api 34$(L) solum-conductor 35$(L) solum-deployer 36$(L) solum-worker* 21:35:58 akshayc: can you try solum app create and then solum assembly create 21:36:18 akshayc: that is good sign.. looks like all solum services are running correctly 21:37:19 akshayc: you are very close to seeing an app deployed in your devstack setup 21:38:05 ok.... i willl try this... where does solum log? i do not see the logs.... 21:38:06 akshayc: you can try that and let us know how it goes 21:38:37 logs are in /opt/stack/logs 21:38:56 akshayc: also you can see info on each service in that services window 21:38:58 akshayc: let us know in #solum if you run into any issues 21:39:15 yep. ok. 21:39:18 cool 21:39:31 is there any other task/blueprint that we should review? 21:39:31 so, for example, to see what is happening with the Solum API service, go to window 32 and look at stdout 21:39:59 ok gpilz 21:40:35 gpilz: you want to update us on your findings about concurrency and wsme/pecan 21:40:39 ? 21:40:45 sure 21:40:53 it isn't an issue 21:41:15 i know this because (a) i wrote a test that confirmed it doesn't happen 21:41:29 cool.. 21:41:39 and ( b ) http://stackoverflow.com/questions/27411102/python-wsme-and-thread-safety 21:41:53 eventually we need to consider adding such concurrency tests to our tempest suite 21:41:54 (that one cost me) 21:42:02 dev - that would be a good idea 21:42:46 mkam: do you know if there are any standard concurrency testing tools being used by OpenStack? 21:42:54 within other OS projects? 21:44:01 I don't know of any that are being used by OpenStack, but I can look into it 21:44:11 that would be nice 21:44:30 thanks mkam 21:44:59 okay are there other blueprints that we want to discuss? 21:45:25 if not I will go to next topic 21:45:32 #topic open discussion 21:46:22 I have an issue I'd like to discuss 21:46:28 sure 21:46:39 but it just applies to me, so we can save it if there is anything else anyone would like to talk about 21:47:24 I guess there are not any major topics on peoples minds 21:47:31 okay - I'm trying to figure out how to get pecan to handle an HTTP PATCH request 21:47:31 so gpilz, floor is yours 21:47:59 if anyone has any experience with anything like this (hacking pecan to support fringe methods) - I'd appreciate the help 21:49:39 in what context are you planning to use patch? 21:49:44 just curious 21:49:56 ironic has similar problem: http://osdir.com/ml/openstack-dev/2014-02/msg02323.html 21:50:24 CAMP specifies the use of HTTP PATCH (with JSON Patch) to update individual attributes of a resource 21:50:38 PATCH allows you to update a resource without using POST with a full REST resource. You could just specify the attributes you want to change. 21:51:02 or PUT 21:52:39 gpilz: I see 21:53:00 gpilz: for time being could PUT work for you? 21:53:21 ravips: how is ironic team resolving this? 21:53:27 the reason PATCH is preferred is it makes it easier for developers to code against 21:53:29 dev - yeah 21:53:37 I'm just trying to flesh out the impl is all 21:53:46 okay 21:53:56 relevant snip "Ironic's API relies on HTTP PATCH to modify resources, which Pecan/WSME does not abstract for us. We're using the python jsonpatch library to parse these requests. I'm not aware of a similar python library for XML support." 21:54:27 thanks ravips 21:54:39 gpilz: that settles it then 21:55:04 dev - actually (re-checking the spec) I *have* to implement PATCH to count as a CAMP impl 21:55:11 I'm sure pecan can do it 21:55:18 it's nothing if not flexible 21:55:24 probably you are right.. 21:55:26 I'm just trying to shave off a day or two 21:55:43 may be hit some experts on pecan in this channel or on the openstack-dev list 21:55:56 thanks 21:57:47 okay thanks everyone for joining today's meeting 21:57:54 #endmeeting