14:01:12 #startmeeting sahara 14:01:14 Meeting started Thu Jan 25 14:01:12 2018 UTC and is due to finish in 60 minutes. The chair is tellesnobrega. Information about MeetBot at http://wiki.debian.org/MeetBot. 14:01:15 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 14:01:18 The meeting name has been set to 'sahara' 14:01:27 o/ 14:01:34 o/ 14:02:09 we can start, tosky is traveling today 14:02:15 #topic News/Updates 14:02:57 I fixed the swift v3 issue, would like you both to take a look, there is a legacy v3/auth there that I didn't have time to look into 14:03:38 if you are ok I think we should merge it and after q3 I will see if we can remove and if needed fix it up 14:04:09 tellesnobrega, v3/auth is working afaik, i was doing proxy users stuff the other day before you put that patch out 14:04:15 other than that, preparing for PTG, and working on file transfer bug that appears to be a pain to solve 14:04:17 mostly working on QA for vanilla upgrade and bug fix. and help a fresher to work on keypair injection. 14:04:34 anyway, my update: several last minute patches about apiv2 from me 14:04:42 jeremyfreudberg, awesome, so it should be kept there 14:04:48 I will test it out here too, 14:05:14 tellesnobrega, yeah v3/auth is ok for now 14:06:19 any more news? 14:06:55 that's it for news, i guess 14:07:01 #topic APIv2 14:07:22 I guess we all have seen a storm of patches from jeremyfreudberg on APIv2 (thanks so much for that) 14:07:51 seems like we are in a great state to finally release APIv2 as experimental and start our process into deprecating api v1 14:08:04 thanks for jeremy's effort 14:08:06 we need it merged today otherwise we need a formal FFE request 14:08:26 yup, with the client path and other final server side tweaks, things are looking good 14:08:32 s/path/patch 14:08:55 since tosky isn't too much around and I think that jeremyfreudberg won't enjoy +2'ing his own patch shuyingya please take some time to review so I can merge it 14:09:13 and we can cut Q3 today with APIv2 finally working 14:09:30 I think we can quickly merge it. If we find problem. just fix it 14:09:39 there are still a few followups for EXPERIMENTAL->CURRENT/SUPPORTED that I want to do in Rocky, but yes, everything is really working now 14:09:43 \o/ 14:09:50 also true 14:09:52 :) 14:10:19 ok, moving on 14:10:31 #topic Vanilla 2.8.2 14:10:46 shuyingya just sent out a patch for vanilla 2.8.2 14:10:51 It works now 14:10:52 what is the status of that 14:10:55 ? 14:11:11 I think the webUI issue can be fix in sahara-extra 14:11:30 https://issues.apache.org/jira/browse/OOZIE-2533 how about apply the workaround patch in sahara-extra? 14:12:00 I don't think remove the java dependency package in sahara side is a good way. 14:13:29 shuyingya, if you think it will work, I like it better than removing java dependency for sure 14:14:02 shuyingya, agree, we can just apply the workaround diff in sahara-extra 14:14:40 get lot of experience in vanilla upgrade. I think we can upgrade Hadoop 3.0 not so far. 14:14:58 shuyingya, that is great to hear 14:15:39 shuyingya: awesome 14:15:49 maybe you could prepare something to get us all the same level of expertise you have now on vanilla upgrade 14:15:53 couple slides maybe 14:15:58 the patch has a UT failed locally as I commented. But I think it is caused by this patch. let's see the CI result 14:16:13 sure 14:16:47 OK. glad to share with you 14:17:23 awesome, let see how it goes and we need to merge it asap (today) 14:18:11 Please review it and I will update it ASAP too. 14:18:16 will do 14:18:29 yes, i will try to review it asap, it's getting late in china 14:18:36 I think it is quite stable now. 14:18:45 awesome 14:19:27 shuyingya, about your unit test 14:19:37 it is failed for me locally too... but that is because there is some pdb statement there 14:19:47 let me try again, removing it 14:20:28 oops :( 14:20:43 running again 14:21:05 yes, still failed 14:21:16 something about the call list is not matching 14:22:24 yes. I remember I use the pdb to trace the error. 14:23:15 forget to remove it. but I think it is weird. I need take a look at mock 14:23:59 ok, i passed it 14:24:03 it's working 14:24:10 jeremyfreudberg, what was wrong? 14:25:02 you can compare patchsets 14:25:04 thanks first, jeremyfreudberg 14:25:21 didn't see a new one 14:25:32 thanks jeremyfreudberg 14:25:32 i just pushed now 14:25:42 shuyingya, that passes locally for me 14:26:05 oops 14:26:13 that makes sense, it does have that call 14:27:01 thanks jeremyfreudberg 14:27:03 awesome, lets wait for the tests and merge it 14:27:45 np 14:27:56 is that it for this topic? (and i will review the rest quickly) 14:28:05 yes 14:28:25 #topic Dublin PTG and Doc day 14:28:53 I was talking with tosky yesterday and he suggested that we do a prep doc day before PTG so we don't waste time thee 14:29:00 what do you think? 14:29:02 I have applied the PTG tsp. let's wait the response. :) 14:29:13 that is awesome shuyingya :) 14:29:21 shuyingya, that is great 14:29:23 hopefully you will be there 14:30:14 That would be great If I can be there. Let's talk about tellesnobrega and tosky's idea 14:30:53 the idea is to save time from PTG 14:30:57 i agree that prep doc day probably makes sense -- what should we be looking for during prep? i mean, most docs are old and weird anyway, what is the priority kind of problem to fix 14:31:26 that is the main idea, we don't know what to look for 14:31:57 the prep day is to get a good look at it and see the main problems so we can get a jump start 14:32:11 right 14:32:16 make sense 14:32:37 what is the best day (weekday) for you both? 14:33:32 mondays and wednesdays are okay 14:33:51 that works for you too shuyingya? 14:34:03 of course coordinating time zone is tricky 14:34:13 I would say anytime is ok, haha 14:34:51 jeremyfreudberg, it is, but we can try to do it in steps, we coordinate parts of doc for each of us and shuyingya can start before us and once we are up we can continue 14:35:15 true 14:35:39 sounds great 14:36:11 cool, I will start organizing it and send out an email on it 14:37:34 moving on 14:37:36 #topic PTL Elections 14:37:56 you probably have seen that PTL election self nomination starts next week 14:38:52 does anyone intend to run this cycle? 14:39:52 i am not in the position time-wise or job-wise to PTL, it is on my mind, probably you can see it happening for T or U release, but not now 14:40:32 jeremyfreudberg, good 14:40:33 tellesnobrega really do a great job on this position. 14:40:40 thanks shuyingya 14:40:50 yes, i hope telles will nominate himself and keep doing good 14:41:00 I will be running since I believe no one else is doing 14:41:05 it 14:41:08 I would be nice if you still nominate yourself :) 14:41:25 but my idea is to get someone else in the position soon as well 14:41:27 +1 14:41:47 so we can take turns, help each other and prepare leadership 14:42:05 SotK, I'm happy to hear that jeremyfreudberg intends to be there soon 14:42:15 I fully support it :) 14:42:46 :) 14:43:16 #topic Open Discussion 14:43:54 https://review.openstack.org/#/c/536002/ and https://review.openstack.org/#/c/537666/ are two of my patches, they really should go into queens-3, so please finish reviewing them 14:44:13 this is more api stuff on the server-side 14:44:31 awesome, looking at the one I didn't look now 14:44:49 both of them were extensively used by me in testing apiv2, so they should work fine 14:44:57 perfect :) 14:45:14 more importantly, tellesnobrega please approve force delete when you can, that definitely needs to go into queens-3 14:45:22 yes 14:45:54 thanks shuyingya for your worfklow+1 s 14:46:19 done, force delete is coming in hot 14:46:21 :) 14:47:08 great 14:47:10 np. jeremyfreudberg. good job 14:47:39 thanks shuyingya 14:47:57 other than those patches i mentioned, i think the only one for queens-3 is vanilla 2.8 14:48:13 yes 14:48:34 i guess we will consider FFE for hbase on sie, shuyingya 14:48:45 yes 14:49:24 shuyingya, not sure if you were connected when i said it, but we will consider FFE for hbase on image 14:50:12 sorry, what's fullname of FFE? 14:50:20 sorry, feature freeze exception 14:50:24 feature freeze exception 14:50:31 if you have time to work on it in the next few weeks, we will help you get it in 14:50:40 if you don't have time, that's okay too 14:51:31 that would be great. I think we can put it off. really busy for downstream tasks 14:51:51 But I promise I will finish it asap 14:52:05 great, thanks again for contributions shuyingya 14:52:11 awesome, thanks shuyingya 14:52:14 np 14:52:28 and we fully understand that balancing upstream and downstream is extremely hard 14:52:58 thanks for your understanding :) 14:53:07 no problem at all 14:54:44 ugh, infra problems on the gate, everything will need recheck later 14:55:00 jeremyfreudberg, I will make sure we get rechecked 14:55:31 "please stand by and don't needlessly recheck job" is the advice for now 14:55:56 sure 14:56:16 will wait as much as I can 14:56:31 so we don't overload infra 14:57:07 btw, i will be offline for a good part of day 14:57:11 unfortunately 14:58:07 no problem 14:58:12 I will be around today the whole dat 14:58:13 day 14:58:22 hopefully I will be able to put out the fires 14:58:25 i have another 30 minutes now, then it will be about 6 hours with minimal interaction 14:58:38 after those 6 hours if the fires are not put out i will help 14:58:54 we are getting very close to meeting time, thanks guys for the hard work during this milestone 14:59:10 yup, thanks all 14:59:28 thanks, see you :) 14:59:56 see you 15:00:01 #endmeeting