| *** gyee has quit IRC | 00:02 | |
| *** chyka has quit IRC | 00:11 | |
| *** jcoufal_ has quit IRC | 00:23 | |
| *** rmascena__ has quit IRC | 00:25 | |
| *** tonyb has quit IRC | 00:33 | |
| *** bobmel has joined #openstack-meeting-alt | 00:33 | |
| *** tonyb has joined #openstack-meeting-alt | 00:35 | |
| *** cloudrancher has quit IRC | 00:36 | |
| *** bobmel has quit IRC | 00:38 | |
| *** spzala has joined #openstack-meeting-alt | 00:49 | |
| *** kumarmn has joined #openstack-meeting-alt | 00:53 | |
| *** hiro-kobayashi has joined #openstack-meeting-alt | 01:09 | |
| *** julim has joined #openstack-meeting-alt | 01:12 | |
| *** spzala has quit IRC | 01:13 | |
| *** julim has quit IRC | 01:14 | |
| *** kumarmn has quit IRC | 01:15 | |
| *** kumarmn has joined #openstack-meeting-alt | 01:20 | |
| *** kumarmn has quit IRC | 01:25 | |
| *** yamahata has quit IRC | 01:26 | |
| *** salv-orl_ has joined #openstack-meeting-alt | 01:28 | |
| *** salv-orlando has quit IRC | 01:31 | |
| *** markstur has quit IRC | 01:31 | |
| *** hongbin_ has joined #openstack-meeting-alt | 01:32 | |
| *** dalgaaf has quit IRC | 01:42 | |
| *** dalgaaf has joined #openstack-meeting-alt | 01:43 | |
| *** spzala has joined #openstack-meeting-alt | 01:52 | |
| *** spzala has quit IRC | 01:57 | |
| *** caowei has joined #openstack-meeting-alt | 02:04 | |
| *** fnaval has quit IRC | 02:22 | |
| *** spzala has joined #openstack-meeting-alt | 02:34 | |
| *** bobmel has joined #openstack-meeting-alt | 02:34 | |
| *** julim has joined #openstack-meeting-alt | 02:35 | |
| *** bobmel has quit IRC | 02:40 | |
| *** spzala has quit IRC | 02:48 | |
| *** yamamoto has quit IRC | 02:53 | |
| *** yamamoto has joined #openstack-meeting-alt | 02:54 | |
| *** yamamoto has quit IRC | 02:59 | |
| *** markstur has joined #openstack-meeting-alt | 03:31 | |
| *** hiro-kobayashi has quit IRC | 03:35 | |
| *** lpetrut has joined #openstack-meeting-alt | 03:39 | |
| *** kumarmn has joined #openstack-meeting-alt | 03:42 | |
| *** janki has joined #openstack-meeting-alt | 03:44 | |
| *** sridharg has joined #openstack-meeting-alt | 03:48 | |
| *** kumarmn has quit IRC | 03:49 | |
| *** shu-mutow-AWAY is now known as shu-mutow | 03:57 | |
| *** hongbin_ has quit IRC | 03:58 | |
| *** ianychoi has quit IRC | 04:02 | |
| *** ianychoi has joined #openstack-meeting-alt | 04:04 | |
| *** kumarmn has joined #openstack-meeting-alt | 04:05 | |
| *** kumarmn has quit IRC | 04:10 | |
| *** lpetrut has quit IRC | 04:11 | |
| *** dklyle has joined #openstack-meeting-alt | 04:17 | |
| *** anilvenkata has joined #openstack-meeting-alt | 04:19 | |
| *** markstur has quit IRC | 04:22 | |
| *** anilvenkata has quit IRC | 04:23 | |
| *** markstur has joined #openstack-meeting-alt | 04:24 | |
| *** ianychoi has quit IRC | 04:28 | |
| *** markvoelker has quit IRC | 04:29 | |
| *** ianychoi has joined #openstack-meeting-alt | 04:31 | |
| *** julim has quit IRC | 04:34 | |
| *** bobmel has joined #openstack-meeting-alt | 04:35 | |
| *** dklyle has quit IRC | 04:37 | |
| *** slaweq has joined #openstack-meeting-alt | 04:39 | |
| *** janki has quit IRC | 04:39 | |
| *** bobmel has quit IRC | 04:40 | |
| *** edmondsw has joined #openstack-meeting-alt | 04:43 | |
| *** edmondsw_ has joined #openstack-meeting-alt | 04:44 | |
| *** slaweq has quit IRC | 04:47 | |
| *** edmondsw has quit IRC | 04:48 | |
| *** lpetrut has joined #openstack-meeting-alt | 04:58 | |
| *** edmondsw_ has quit IRC | 04:59 | |
| *** lpetrut has quit IRC | 05:27 | |
| *** bobmel has joined #openstack-meeting-alt | 05:36 | |
| *** marios has joined #openstack-meeting-alt | 05:39 | |
| *** bobmel has quit IRC | 05:41 | |
| *** edmondsw has joined #openstack-meeting-alt | 05:41 | |
| *** edmondsw has quit IRC | 05:45 | |
| *** janki has joined #openstack-meeting-alt | 05:49 | |
| *** shu-mutow is now known as shu-mutow-AWAY | 05:59 | |
| *** edmondsw has joined #openstack-meeting-alt | 06:02 | |
| *** edmondsw has quit IRC | 06:06 | |
| *** MarkBaker has quit IRC | 06:07 | |
| *** julim has joined #openstack-meeting-alt | 06:08 | |
| *** links has joined #openstack-meeting-alt | 06:16 | |
| *** bhavik1 has joined #openstack-meeting-alt | 06:17 | |
| *** bhavik1 has quit IRC | 06:23 | |
| *** armaan has joined #openstack-meeting-alt | 06:24 | |
| *** armaan has quit IRC | 06:24 | |
| *** trinaths has joined #openstack-meeting-alt | 06:24 | |
| *** jtomasek has joined #openstack-meeting-alt | 06:25 | |
| *** lpetrut has joined #openstack-meeting-alt | 06:29 | |
| *** markvoelker has joined #openstack-meeting-alt | 06:30 | |
| *** MarkBaker has joined #openstack-meeting-alt | 06:31 | |
| *** slaweq has joined #openstack-meeting-alt | 06:32 | |
| *** kopecmartin has joined #openstack-meeting-alt | 06:33 | |
| *** trungnv has joined #openstack-meeting-alt | 06:41 | |
| *** slaweq has quit IRC | 06:42 | |
| *** salv-orl_ has quit IRC | 06:45 | |
| *** salv-orlando has joined #openstack-meeting-alt | 06:46 | |
| *** tesseract has joined #openstack-meeting-alt | 06:49 | |
| *** salv-orlando has quit IRC | 06:50 | |
| *** markstur has quit IRC | 06:55 | |
| *** ianychoi has quit IRC | 07:00 | |
| *** salv-orlando has joined #openstack-meeting-alt | 07:00 | |
| *** markvoelker has quit IRC | 07:04 | |
| *** armaan has joined #openstack-meeting-alt | 07:06 | |
| *** ianychoi has joined #openstack-meeting-alt | 07:11 | |
| *** slaweq has joined #openstack-meeting-alt | 07:16 | |
| *** rcernin has quit IRC | 07:16 | |
| *** jchhatbar has joined #openstack-meeting-alt | 07:21 | |
| *** oikiki has joined #openstack-meeting-alt | 07:23 | |
| *** janki has quit IRC | 07:23 | |
| *** edmondsw has joined #openstack-meeting-alt | 07:24 | |
| *** armaan has quit IRC | 07:28 | |
| *** edmondsw has quit IRC | 07:28 | |
| *** rmart04 has joined #openstack-meeting-alt | 07:31 | |
| *** bobmel has joined #openstack-meeting-alt | 07:37 | |
| *** bobmel has quit IRC | 07:42 | |
| *** derekjhyang has quit IRC | 07:44 | |
| *** edmondsw has joined #openstack-meeting-alt | 07:45 | |
| *** ethfci has quit IRC | 07:45 | |
| *** tssurya has joined #openstack-meeting-alt | 07:45 | |
| *** ethfci has joined #openstack-meeting-alt | 07:45 | |
| *** macermak has joined #openstack-meeting-alt | 07:46 | |
| *** ethfci has quit IRC | 07:47 | |
| *** ethfci has joined #openstack-meeting-alt | 07:47 | |
| *** edmondsw has quit IRC | 07:49 | |
| *** dkehn_ has joined #openstack-meeting-alt | 07:50 | |
| *** spzala has joined #openstack-meeting-alt | 07:53 | |
| *** dkehn has quit IRC | 07:54 | |
| *** spzala has quit IRC | 07:57 | |
| *** markvoelker has joined #openstack-meeting-alt | 08:01 | |
| *** oikiki has quit IRC | 08:03 | |
| *** edmondsw has joined #openstack-meeting-alt | 08:05 | |
| *** kumarmn has joined #openstack-meeting-alt | 08:06 | |
| *** armaan has joined #openstack-meeting-alt | 08:08 | |
| *** edmondsw has quit IRC | 08:09 | |
| *** kumarmn has quit IRC | 08:11 | |
| *** vds has quit IRC | 08:18 | |
| *** dsariel has joined #openstack-meeting-alt | 08:20 | |
| *** priteau has joined #openstack-meeting-alt | 08:25 | |
| *** edmondsw has joined #openstack-meeting-alt | 08:26 | |
| *** bfernando has joined #openstack-meeting-alt | 08:27 | |
| *** edmondsw has quit IRC | 08:30 | |
| *** markvoelker has quit IRC | 08:34 | |
| *** derekh has joined #openstack-meeting-alt | 08:41 | |
| *** MarkBaker has quit IRC | 08:42 | |
| *** hiro-kobayashi has joined #openstack-meeting-alt | 08:42 | |
| *** radeks has joined #openstack-meeting-alt | 08:48 | |
| *** ianw is now known as ianw_pto | 08:54 | |
| *** armaan has quit IRC | 08:58 | |
| *** armaan has joined #openstack-meeting-alt | 08:58 | |
| *** masahito has joined #openstack-meeting-alt | 09:00 | |
| masahito | #startmeeting blazar | 09:01 |
|---|---|---|
| openstack | Meeting started Tue Apr 24 09:01:06 2018 UTC and is due to finish in 60 minutes. The chair is masahito. Information about MeetBot at http://wiki.debian.org/MeetBot. | 09:01 |
| openstack | Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. | 09:01 |
| *** openstack changes topic to " (Meeting topic: blazar)" | 09:01 | |
| openstack | The meeting name has been set to 'blazar' | 09:01 |
| masahito | hi all, time to blazar meeting | 09:01 |
| masahito | #topic RollCall | 09:01 |
| *** openstack changes topic to "RollCall (Meeting topic: blazar)" | 09:01 | |
| hiro-kobayashi | o/ | 09:02 |
| masahito | Todya's agenda is | 09:02 |
| masahito | 1. September PTG | 09:02 |
| masahito | 2. PlacementAPI support | 09:02 |
| masahito | 3. Reservation re-allocation | 09:02 |
| masahito | 4. AOB | 09:02 |
| masahito | anything else? | 09:02 |
| priteau | o/ | 09:02 |
| masahito | hiro-kobayashi, priteau: hi | 09:02 |
| priteau | Good morning masahito and hiro-kobayashi | 09:02 |
| masahito | priteau: good morning :-) | 09:02 |
| *** armaan has quit IRC | 09:03 | |
| hiro-kobayashi | hi priteau | 09:03 |
| masahito | bertys and GeraldK told me they can't join today's meeting. | 09:03 |
| masahito | #topic September PTG survey | 09:03 |
| *** openstack changes topic to "September PTG survey (Meeting topic: blazar)" | 09:03 | |
| masahito | The official announced the next PTG for Stein will be held in Denver on 10th-14th Sep. | 09:05 |
| masahito | http://lists.openstack.org/pipermail/openstack-dev/2018-April/129564.html | 09:05 |
| masahito | They asked me that we plan to have PTG during this week. | 09:06 |
| masahito | Do you both have a plan to join the PTG? | 09:07 |
| hiro-kobayashi | Yes, I plan to join it. | 09:07 |
| priteau | I can't say yet, it will depend if my employer is happy for me to travel there | 09:07 |
| masahito | They also requested me replaying whether to have or not to have until 2nd May. | 09:08 |
| masahito | I plan to join it, too. | 09:09 |
| priteau | If it's just hiro-kobayashi and you, would you still hold a Blazar meeting there? | 09:09 |
| -masahito- So looks like few of us will attend the PTG. I'll replay we plan to have meeting. | 09:10 | |
| hiro-kobayashi | I hope new people will join it. | 09:10 |
| priteau | That would be great. | 09:10 |
| masahito | priteau: If it's just us, we can implement new features there instead of discussion. | 09:11 |
| priteau | A hackathon? :-) | 09:11 |
| masahito | yes :-) | 09:11 |
| hiro-kobayashi | Sounds good :-) | 09:12 |
| priteau | I will try to let you know if I can attend as soon as possible, but I might know only weeks from now | 09:12 |
| masahito | I plan to join other project meeting. So I'll be there in either case. | 09:12 |
| masahito | priteau: got it. The early bird ticket will end by 10th May. | 09:13 |
| priteau | I will get a ticket anyway, since refund is possible until end of August | 09:14 |
| *** sridharg has quit IRC | 09:14 | |
| masahito | sounds nice. | 09:14 |
| masahito | anything else for this topic? | 09:15 |
| masahito | #topic PlacementAPI support | 09:16 |
| *** openstack changes topic to "PlacementAPI support (Meeting topic: blazar)" | 09:16 | |
| masahito | I heard hiro-kobayashi has something to discuss for the topic. | 09:17 |
| hiro-kobayashi | yes | 09:17 |
| hiro-kobayashi | I created an etherpad for discussing placement support: https://etherpad.openstack.org/p/blazar-placement-design | 09:17 |
| hiro-kobayashi | I wrote requirements and basic approaches down there. | 09:18 |
| *** sridharg has joined #openstack-meeting-alt | 09:18 | |
| hiro-kobayashi | The approach 1, 3 and 4 were basically copied from the Dublin PTG etherpad. | 09:18 |
| hiro-kobayashi | I don't understand much about approach 4, though. | 09:19 |
| priteau | Thank you hiro-kobayashi for creating the document | 09:19 |
| priteau | I think there is also the concept of aggregates in placement API | 09:20 |
| masahito | hiro-kobayashi: Nice works. | 09:20 |
| hiro-kobayashi | priteau: Yes it would be. I appreciate it if you could write your idea down there | 09:20 |
| hiro-kobayashi | And, feel free to edit existing ideas | 09:21 |
| hiro-kobayashi | I think it's better to make a consensus on the basic approach in the Blazar team before Vancouver summit. | 09:21 |
| priteau | I will review and add more details as soon as possible, hopefully this week | 09:23 |
| masahito | I'd like to add some items for each approach. The way of user interaction with Nova will be changed, etc. | 09:23 |
| hiro-kobayashi | Thanks priteau! | 09:23 |
| hiro-kobayashi | Thanks! | 09:23 |
| masahito | Me too. I'll add my comments, too. | 09:23 |
| masahito | After one or two weeks later, we should share this to Placement team. | 09:24 |
| masahito | As hiro-kobayashi said, that should be done before the Vancouver summit. | 09:24 |
| hiro-kobayashi | masahito: Agree. | 09:25 |
| *** panda|bbl is now known as panda|ruck | 09:25 | |
| masahito | anything else? | 09:26 |
| hiro-kobayashi | Nothing from me. | 09:27 |
| masahito | #topic Reservation re-allocation | 09:27 |
| *** openstack changes topic to "Reservation re-allocation (Meeting topic: blazar)" | 09:27 | |
| masahito | I'm working on this spec now. https://docs.openstack.org/blazar/latest/specs/queens/flavors-extra-specs.html | 09:28 |
| masahito | I want talk about the third work items, Add re-allocation logic to ComputeHostExtraCapability management. | 09:28 |
| masahito | Based on the discussion on gerrit, Blazar needs to re-allocate reservation once the extra_capability is updated, changed or deleted. | 09:30 |
| masahito | However, Host delete API doesn't allow cloud admins to remove hosts if there are reservations to use the hosts. | 09:31 |
| *** markvoelker has joined #openstack-meeting-alt | 09:31 | |
| masahito | IMHO, if the Update Host API follows the way of the Delete Host API, the Blazar shouldn't accept the update itself. | 09:32 |
| priteau | One way to resolve the problem would be to prevent extra capability update if the node is reserved | 09:32 |
| hiro-kobayashi | priteau: +1 | 09:33 |
| priteau | However, that is quite a strong requirement. A better way to handle it would be to prevent extra capability update only if the node has reservations that are using resource_properties | 09:33 |
| masahito | make sense. | 09:34 |
| priteau | But we could start with the first approach, and then make it smarter | 09:34 |
| hiro-kobayashi | Good approach | 09:34 |
| masahito | Alternative is disabling the host and resource-monitoring re-allocate the reservations. | 09:35 |
| priteau | Will it re-allocate a running reservation? I don't remember the behavior | 09:36 |
| masahito | When the reservation is active, resource_changed flag becomes true at least. | 09:37 |
| hiro-kobayashi | Currently It is not allowed to reallocate if it's active. | 09:37 |
| *** bobmel has joined #openstack-meeting-alt | 09:38 | |
| hiro-kobayashi | masahito: I think blazar should not allow reallocate if it's active except a failure recovery situation. | 09:38 |
| masahito | hiro-kobayashi: IIRC, in addition to resource_changed flag, the missing_resource flag becomes true if there is no alternative resource. | 09:38 |
| hiro-kobayashi | yes | 09:39 |
| masahito | hiro-kobayashi: because it raises complexity of Blazar? | 09:40 |
| hiro-kobayashi | masahito: yes. | 09:41 |
| masahito | got it. | 09:42 |
| *** bobmel has quit IRC | 09:42 | |
| masahito | That's from my side. any comment for the topic? | 09:42 |
| hiro-kobayashi | If blazar allow reallocation of active reservation, blazar has to handle running instances on removed hosts. | 09:43 |
| priteau | masahito: I would prefer starting with a simple approach first and increasing complexity later | 09:44 |
| hiro-kobayashi | How about alining with host reservation update? | 09:45 |
| hiro-kobayashi | https://github.com/openstack/blazar/blob/master/blazar/plugins/oshosts/host_plugin.py#L527 | 09:45 |
| masahito | I agreed priteau's idea. I plan to start with your suggestion first. | 09:45 |
| masahito | I'd like to talk the goal of the BP. | 09:46 |
| masahito | If the reservation is in pending status, Blazar try to re-allocate. If it's in active, it fails. right? | 09:48 |
| masahito | Even if the status is in pending, the update is not allowed in case of missing_resource is true. | 09:48 |
| hiro-kobayashi | I think that's right. | 09:49 |
| masahito | Make sense. Anyway I'll start with the simple way to do this. | 09:51 |
| masahito | #topic AOB | 09:51 |
| *** openstack changes topic to "AOB (Meeting topic: blazar)" | 09:51 | |
| masahito | Does someone have something to share/discuss? | 09:51 |
| masahito | last 10 mins | 09:51 |
| hiro-kobayashi | Nothing from me | 09:52 |
| priteau | I have completed the support for Ironic using Jay's unmerged Nova aggregates patch. The changes to Blazar were quite small. | 09:53 |
| masahito | priteau: Great work! | 09:53 |
| hiro-kobayashi | Good news! | 09:54 |
| priteau | Moving to placement API will probably be a lot more work ;-) | 09:55 |
| masahito | Yes. But that's a really good news. We can move things forward. | 09:56 |
| masahito | priteau: btw, do you have some specific topics in next meeting now? The day of next meeting, 1st May., is holiday in Japan. | 09:58 |
| priteau | Nothing specific from me. | 09:58 |
| priteau | Would you like to skip the meeting? | 09:58 |
| masahito | If nothing and bertys and GeraldK also doesn't have nothing, I want to skip the meeting. | 09:58 |
| masahito | priteau: yes. | 09:58 |
| priteau | No problem | 09:59 |
| masahito | I'll ask them later. If they doesn't have no topics, I announce the skipping in openstack-dev ML. | 09:59 |
| -masahito- thanks. | 10:00 | |
| *** julim has quit IRC | 10:00 | |
| masahito | Running out of time. Nice discussion and greate updates! | 10:01 |
| masahito | Thanks all. bye. | 10:01 |
| priteau | Bye everyone, thank you! | 10:01 |
| masahito | #endmeeting | 10:01 |
| *** openstack changes topic to "OpenStack Meetings || https://wiki.openstack.org/wiki/Meetings/" | 10:01 | |
| openstack | Meeting ended Tue Apr 24 10:01:46 2018 UTC. Information about MeetBot at http://wiki.debian.org/MeetBot . (v 0.1.4) | 10:01 |
| openstack | Minutes: http://eavesdrop.openstack.org/meetings/blazar/2018/blazar.2018-04-24-09.01.html | 10:01 |
| openstack | Minutes (text): http://eavesdrop.openstack.org/meetings/blazar/2018/blazar.2018-04-24-09.01.txt | 10:01 |
| openstack | Log: http://eavesdrop.openstack.org/meetings/blazar/2018/blazar.2018-04-24-09.01.log.html | 10:01 |
| hiro-kobayashi | Thank you all! bye! | 10:01 |
| *** hiro-kobayashi has quit IRC | 10:01 | |
| *** armaan has joined #openstack-meeting-alt | 10:02 | |
| *** trungnv has quit IRC | 10:02 | |
| *** rmart04_ has joined #openstack-meeting-alt | 10:04 | |
| *** macermak has quit IRC | 10:05 | |
| *** macermak has joined #openstack-meeting-alt | 10:05 | |
| *** markvoelker has quit IRC | 10:06 | |
| *** rmart04 has quit IRC | 10:06 | |
| *** rmart04_ is now known as rmart04 | 10:06 | |
| *** bobmel has joined #openstack-meeting-alt | 10:07 | |
| *** sambetts_ is now known as sambetts | 10:11 | |
| *** priteau has quit IRC | 10:11 | |
| *** priteau has joined #openstack-meeting-alt | 10:11 | |
| *** rmart04 has quit IRC | 10:17 | |
| *** bfernando has quit IRC | 10:25 | |
| *** armaan has quit IRC | 10:28 | |
| *** armaan has joined #openstack-meeting-alt | 10:29 | |
| *** vds has joined #openstack-meeting-alt | 10:31 | |
| *** armaan has quit IRC | 10:33 | |
| *** rmart04 has joined #openstack-meeting-alt | 10:33 | |
| *** armaan has joined #openstack-meeting-alt | 10:33 | |
| *** vds has quit IRC | 10:38 | |
| *** rmart04 has quit IRC | 10:46 | |
| *** macermak has quit IRC | 10:57 | |
| *** markvoelker has joined #openstack-meeting-alt | 11:02 | |
| *** chyka has joined #openstack-meeting-alt | 11:06 | |
| *** jcoufal has joined #openstack-meeting-alt | 11:07 | |
| *** rmart04 has joined #openstack-meeting-alt | 11:09 | |
| *** chyka has quit IRC | 11:11 | |
| *** caowei has quit IRC | 11:13 | |
| *** trinaths has quit IRC | 11:14 | |
| *** masahito has quit IRC | 11:15 | |
| *** pbourke has quit IRC | 11:18 | |
| *** pbourke has joined #openstack-meeting-alt | 11:18 | |
| *** armaan has quit IRC | 11:26 | |
| *** armaan has joined #openstack-meeting-alt | 11:26 | |
| *** macermak has joined #openstack-meeting-alt | 11:29 | |
| *** armaan has quit IRC | 11:33 | |
| *** armaan has joined #openstack-meeting-alt | 11:33 | |
| *** macermak_ has joined #openstack-meeting-alt | 11:35 | |
| *** markvoelker has quit IRC | 11:36 | |
| *** macermak has quit IRC | 11:38 | |
| *** macermak_ has quit IRC | 11:45 | |
| *** spzala has joined #openstack-meeting-alt | 11:53 | |
| *** liuyulong has joined #openstack-meeting-alt | 11:54 | |
| *** spzala has quit IRC | 11:57 | |
| *** bfernando has joined #openstack-meeting-alt | 12:00 | |
| *** dlenrow has quit IRC | 12:05 | |
| *** dlenrow has joined #openstack-meeting-alt | 12:06 | |
| *** raildo has joined #openstack-meeting-alt | 12:08 | |
| *** zhongjun_ has quit IRC | 12:09 | |
| *** zhongjun_ has joined #openstack-meeting-alt | 12:09 | |
| *** matrohon has joined #openstack-meeting-alt | 12:12 | |
| *** yamamoto has joined #openstack-meeting-alt | 12:14 | |
| *** efried has joined #openstack-meeting-alt | 12:16 | |
| *** markvoelker has joined #openstack-meeting-alt | 12:16 | |
| *** edmondsw has joined #openstack-meeting-alt | 12:28 | |
| *** lbragstad has joined #openstack-meeting-alt | 12:29 | |
| *** edmondsw_ has joined #openstack-meeting-alt | 12:29 | |
| *** tpsilva has joined #openstack-meeting-alt | 12:29 | |
| *** edmondsw has quit IRC | 12:33 | |
| *** vgreen has joined #openstack-meeting-alt | 12:34 | |
| *** panda|ruck is now known as panda|ruck|lunch | 12:37 | |
| *** edmondsw_ has quit IRC | 12:44 | |
| *** edmondsw has joined #openstack-meeting-alt | 12:48 | |
| *** felipemonteiro__ has joined #openstack-meeting-alt | 12:52 | |
| *** edmondsw has quit IRC | 12:52 | |
| *** spzala has joined #openstack-meeting-alt | 12:53 | |
| *** macermak has joined #openstack-meeting-alt | 12:54 | |
| *** learnmore has joined #openstack-meeting-alt | 12:55 | |
| *** spzala has quit IRC | 12:57 | |
| *** edmondsw has joined #openstack-meeting-alt | 12:59 | |
| *** fuqiao has joined #openstack-meeting-alt | 13:00 | |
| *** edmondsw has quit IRC | 13:01 | |
| *** alexchadin has joined #openstack-meeting-alt | 13:02 | |
| *** cloudrancher has joined #openstack-meeting-alt | 13:04 | |
| *** learnmore has quit IRC | 13:08 | |
| *** AlanClark has joined #openstack-meeting-alt | 13:11 | |
| *** armaan has quit IRC | 13:12 | |
| *** armaan has joined #openstack-meeting-alt | 13:13 | |
| *** kumarmn has joined #openstack-meeting-alt | 13:14 | |
| *** salv-orlando has quit IRC | 13:18 | |
| *** salv-orlando has joined #openstack-meeting-alt | 13:19 | |
| *** armaan has quit IRC | 13:20 | |
| *** armaan has joined #openstack-meeting-alt | 13:21 | |
| *** dave-mccowan has joined #openstack-meeting-alt | 13:22 | |
| *** salv-orlando has quit IRC | 13:24 | |
| *** spzala has joined #openstack-meeting-alt | 13:26 | |
| *** edmondsw has joined #openstack-meeting-alt | 13:27 | |
| *** liuyulong has quit IRC | 13:27 | |
| *** yamamoto has quit IRC | 13:28 | |
| *** yamamoto has joined #openstack-meeting-alt | 13:28 | |
| *** edmondsw has quit IRC | 13:31 | |
| *** armaan has quit IRC | 13:32 | |
| *** fuqiao has quit IRC | 13:32 | |
| *** armaan has joined #openstack-meeting-alt | 13:32 | |
| *** fuqiao has joined #openstack-meeting-alt | 13:32 | |
| *** vds has joined #openstack-meeting-alt | 13:36 | |
| *** guyr-infinidat has joined #openstack-meeting-alt | 13:36 | |
| *** panda|ruck|lunch is now known as panda|ruck | 13:43 | |
| *** alexchadin has quit IRC | 13:50 | |
| *** armaan has quit IRC | 13:51 | |
| *** armaan has joined #openstack-meeting-alt | 13:51 | |
| *** felipemonteiro__ has quit IRC | 13:57 | |
| *** felipemonteiro__ has joined #openstack-meeting-alt | 13:57 | |
| *** jaypipes has quit IRC | 13:58 | |
| *** jaypipes has joined #openstack-meeting-alt | 13:58 | |
| *** hongbin_ has joined #openstack-meeting-alt | 13:59 | |
| *** edmondsw has joined #openstack-meeting-alt | 13:59 | |
| *** dustins has joined #openstack-meeting-alt | 14:00 | |
| *** amito has quit IRC | 14:02 | |
| *** amito has joined #openstack-meeting-alt | 14:02 | |
| *** edmondsw has quit IRC | 14:04 | |
| *** yamahata has joined #openstack-meeting-alt | 14:04 | |
| *** edmondsw has joined #openstack-meeting-alt | 14:04 | |
| *** edmondsw has quit IRC | 14:09 | |
| *** edmondsw has joined #openstack-meeting-alt | 14:09 | |
| *** matrohon has quit IRC | 14:11 | |
| *** edmondsw has quit IRC | 14:14 | |
| *** iyamahat has joined #openstack-meeting-alt | 14:19 | |
| *** zhongjun_ has quit IRC | 14:19 | |
| *** cloudrancher has quit IRC | 14:24 | |
| *** alexchadin has joined #openstack-meeting-alt | 14:25 | |
| *** iyamahat has quit IRC | 14:26 | |
| *** iyamahat has joined #openstack-meeting-alt | 14:26 | |
| *** edmondsw has joined #openstack-meeting-alt | 14:35 | |
| *** namnh has joined #openstack-meeting-alt | 14:35 | |
| *** edmondsw has quit IRC | 14:40 | |
| *** links has quit IRC | 14:42 | |
| *** rossella_s has quit IRC | 14:42 | |
| *** edmondsw has joined #openstack-meeting-alt | 14:42 | |
| *** gagehugo has joined #openstack-meeting-alt | 14:46 | |
| *** edmondsw has quit IRC | 14:46 | |
| *** alexchadin has quit IRC | 14:48 | |
| *** markstur has joined #openstack-meeting-alt | 14:57 | |
| *** yamamoto has quit IRC | 14:58 | |
| *** rossella_s has joined #openstack-meeting-alt | 14:59 | |
| *** rmart04 has quit IRC | 15:01 | |
| *** wxy| has joined #openstack-meeting-alt | 15:02 | |
| *** markstur has quit IRC | 15:02 | |
| *** tesseract has quit IRC | 15:06 | |
| *** tesseract has joined #openstack-meeting-alt | 15:06 | |
| *** cloudrancher has joined #openstack-meeting-alt | 15:10 | |
| *** macermak has quit IRC | 15:12 | |
| *** felipemonteiro_ has joined #openstack-meeting-alt | 15:13 | |
| *** alexchadin has joined #openstack-meeting-alt | 15:13 | |
| *** armaan has quit IRC | 15:14 | |
| *** armaan has joined #openstack-meeting-alt | 15:14 | |
| *** felipemonteiro__ has quit IRC | 15:16 | |
| *** kumarmn has quit IRC | 15:25 | |
| *** dklyle has joined #openstack-meeting-alt | 15:29 | |
| *** cloudrancher has quit IRC | 15:31 | |
| *** edmondsw has joined #openstack-meeting-alt | 15:33 | |
| *** salv-orl_ has joined #openstack-meeting-alt | 15:35 | |
| *** kumarmn has joined #openstack-meeting-alt | 15:37 | |
| *** gyee has joined #openstack-meeting-alt | 15:38 | |
| *** kumarmn has quit IRC | 15:40 | |
| *** cloudrancher has joined #openstack-meeting-alt | 15:41 | |
| *** cloudrancher has quit IRC | 15:42 | |
| *** kumarmn has joined #openstack-meeting-alt | 15:43 | |
| *** kumarmn has quit IRC | 15:45 | |
| *** alexchadin has quit IRC | 15:46 | |
| *** matrohon has joined #openstack-meeting-alt | 15:47 | |
| *** kopecmartin has quit IRC | 15:48 | |
| *** felipemonteiro_ has quit IRC | 15:49 | |
| *** felipemonteiro_ has joined #openstack-meeting-alt | 15:49 | |
| *** chyka has joined #openstack-meeting-alt | 15:50 | |
| *** trinaths has joined #openstack-meeting-alt | 15:52 | |
| *** chyka_ has joined #openstack-meeting-alt | 15:53 | |
| *** jbadiapa has quit IRC | 15:55 | |
| *** chyka has quit IRC | 15:56 | |
| *** rossella_s has quit IRC | 15:59 | |
| *** yamamoto has joined #openstack-meeting-alt | 15:59 | |
| *** rossella_s has joined #openstack-meeting-alt | 15:59 | |
| lbragstad | #startmeeting keystone | 16:00 |
| openstack | Meeting started Tue Apr 24 16:00:10 2018 UTC and is due to finish in 60 minutes. The chair is lbragstad. Information about MeetBot at http://wiki.debian.org/MeetBot. | 16:00 |
| openstack | Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. | 16:00 |
| lbragstad | ping ayoung, breton, cmurphy, dstanek, gagehugo, henrynash, hrybacki, knikolla, lamt, lbragstad, lwanderley, kmalloc, rderose, rodrigods, samueldmq, spilla, aselius, dpar, jdennis, ruan_he, wxy, sonuk | 16:00 |
| *** openstack changes topic to " (Meeting topic: keystone)" | 16:00 | |
| openstack | The meeting name has been set to 'keystone' | 16:00 |
| gagehugo | o/ | 16:00 |
| hrybacki | o/ | 16:00 |
| lbragstad | #link https://etherpad.openstack.org/p/keystone-weekly-meeting | 16:00 |
| lbragstad | agenda ^ | 16:00 |
| cmurphy | o/ | 16:00 |
| knikolla | o/ | 16:00 |
| wxy| | o/ | 16:00 |
| kmalloc | o/ | 16:01 |
| lbragstad | #topic specifications | 16:01 |
| *** openstack changes topic to "specifications (Meeting topic: keystone)" | 16:01 | |
| lbragstad | just a heads up - we're now past the specification proposal deadline | 16:02 |
| lbragstad | we'll continue to press on with reviews of what's been proposed | 16:02 |
| *** spilla has joined #openstack-meeting-alt | 16:02 | |
| lbragstad | next important deadline will be specification freeze which is rocky-2, or June 8th | 16:03 |
| lbragstad | #topic consul | 16:03 |
| *** openstack changes topic to "consul (Meeting topic: keystone)" | 16:03 | |
| lbragstad | #link http://eavesdrop.openstack.org/irclogs/%23openstack-tc/%23openstack-tc.2018-04-18.log.html#t2018-04-18T13:21:23 | 16:03 |
| lbragstad | ^ consul came up again in the -tc channel | 16:03 |
| lbragstad | which has been a topic that comes up every once in a while | 16:03 |
| hrybacki | 'discussed but never implemented' -- just always a low prio? | 16:04 |
| *** armaan has quit IRC | 16:04 | |
| *** pbourke has quit IRC | 16:04 | |
| lbragstad | i'm not even sure we made it that far | 16:04 |
| *** pbourke has joined #openstack-meeting-alt | 16:04 | |
| *** yamamoto has quit IRC | 16:04 | |
| lbragstad | here is a list of every time it's been mentioned in keystone http://paste.openstack.org/raw/719483/ (i think) | 16:04 |
| *** armaan has joined #openstack-meeting-alt | 16:05 | |
| lbragstad | list of logs* | 16:05 |
| gagehugo | hmm | 16:05 |
| hrybacki | nice tool lbragstad | 16:05 |
| lbragstad | does this seem like something that would be appropriate for the idea log? http://specs.openstack.org/openstack/keystone-specs/specs/keystone/ideas/README.html | 16:05 |
| *** kumarmn has joined #openstack-meeting-alt | 16:06 | |
| cmurphy | what is the use case? | 16:06 |
| lbragstad | so - consul is service discovery software | 16:06 |
| lbragstad | the use case would be that we would off load service catalog stuff to consul instead of reinventing the wheel | 16:06 |
| lbragstad | but kmalloc knows way more about it than i do | 16:07 |
| kmalloc | if we want something clever, like services self-registering. | 16:07 |
| kmalloc | consul is great | 16:07 |
| kmalloc | otherwise it's not much different than we have now | 16:07 |
| *** armaan_ has joined #openstack-meeting-alt | 16:07 | |
| cmurphy | so rewrite the service catalog in order to avoid writing a new service catalog? | 16:07 |
| cmurphy | sorry didn't mean for that to come out so cynical | 16:08 |
| lbragstad | i was under the impression we would implement a driver or work directly with services to leverage consul | 16:08 |
| lbragstad | s/driver/catalog driver/ | 16:08 |
| *** armaan has quit IRC | 16:08 | |
| *** kumarmn_ has joined #openstack-meeting-alt | 16:09 | |
| *** armaan_ has quit IRC | 16:09 | |
| *** bfernando has quit IRC | 16:10 | |
| lbragstad | (i believe the second option would offload the responsibility of the catalog to a separate service, and users would talk to consul to figure out where to go to interact with services in a deployment) | 16:10 |
| *** kumarmn has quit IRC | 16:10 | |
| *** iyamahat has quit IRC | 16:10 | |
| *** rossella_s has quit IRC | 16:10 | |
| *** fuqiao has quit IRC | 16:12 | |
| *** armaan has joined #openstack-meeting-alt | 16:12 | |
| kmalloc | lbragstad: coreect | 16:13 |
| *** armaan has quit IRC | 16:13 | |
| kmalloc | if we are doing that | 16:13 |
| kmalloc | but... i am against the speculative "maybe" bit of a driver | 16:13 |
| kmalloc | if consul isn't really a dep of openstack, I don't want to add it | 16:13 |
| cmurphy | ++ | 16:13 |
| kmalloc | it's why i pushed for picking zookeeper...or consul...or even etcd | 16:13 |
| kmalloc | and we ended up with "well ALL OF THEM/any" | 16:14 |
| cmurphy | etcd is a base service of openstack | 16:14 |
| lbragstad | ^ that's a good point that someone brought up in the discussion a couple days ago | 16:14 |
| kmalloc | etcd is mostly a base service. | 16:14 |
| kmalloc | so, if we are picking one, i'd say hard set on etcd | 16:14 |
| kmalloc | not "add consul" | 16:14 |
| kmalloc | and work around etcd's limitations | 16:14 |
| kmalloc | if we can | 16:14 |
| *** rossella_s has joined #openstack-meeting-alt | 16:15 | |
| *** yamahata has quit IRC | 16:15 | |
| lbragstad | ok - the fact that consul gets brought up every few/six months tells me it's worth investigating, but i don't expect anyone to sign up for this | 16:15 |
| lbragstad | that said, where should we document this? | 16:15 |
| kmalloc | *shrug* | 16:16 |
| lbragstad | it feels like a PoC of some kind is needed | 16:16 |
| kmalloc | i'm going to be very anti-consul for catalog unless we have some serious weight behind it... even with a PoC | 16:16 |
| cmurphy | i'd like to see some evidence or arguments for why this is better than what we have now and worth us putting in the effort, or else see someone with a specific use case step up to do the work | 16:16 |
| cmurphy | we don't have the bandwidth to take on projects just because | 16:16 |
| kmalloc | cmurphy: ++ | 16:16 |
| gagehugo | yeah.. | 16:16 |
| lbragstad | oh - sorry, i meant in more of a general sense and not just consul | 16:17 |
| wxy| | So users want Keystone to support service self-registering One of the driver may be consul. | 16:17 |
| wxy| | right? | 16:17 |
| kmalloc | wxy|: sortof... | 16:17 |
| kmalloc | self-registering and direct query by endpoints vs. in token | 16:17 |
| lbragstad | cmurphy: i completely agree about bandwidth | 16:18 |
| lbragstad | i don't want to propose more work for the sake of work, but during the tc discussion it was apparent that this particular thing has been brought up consistently since ~2015 | 16:21 |
| lbragstad | i guess my end goal was to either open a wishlist bug describing the thing that gets brought up or somehow make it discoverable so that if someone did want to do a PoC they wouldn't have to dig through all the context again | 16:22 |
| *** tssurya has quit IRC | 16:23 | |
| hrybacki | I think adding this to the idea list and letting relevant ideas formulate there until a strong ask arises is a good idea | 16:24 |
| hrybacki | then we have some documented history and can make a more educated decision when priorities begin to shift? | 16:24 |
| lbragstad | if people don't have any objections to that, i can take a stab at proposing it to the idea log | 16:24 |
| cmurphy | i think that's fine, as long as the idea is "do some research on this" and not "implement this" | 16:25 |
| lbragstad | ++ | 16:25 |
| lbragstad | agreed | 16:25 |
| hrybacki | ++ | 16:25 |
| lbragstad | #action lbragstad to fill out and idea form for consul | 16:25 |
| *** jessegler has joined #openstack-meeting-alt | 16:25 | |
| *** bhavik1 has joined #openstack-meeting-alt | 16:25 | |
| lbragstad | #topic idempotent migrations | 16:25 |
| *** openstack changes topic to "idempotent migrations (Meeting topic: keystone)" | 16:26 | |
| lbragstad | hrybacki: o/ | 16:26 |
| hrybacki | o/ | 16:26 |
| hrybacki | So stemming from the review posted by Ade | 16:26 |
| hrybacki | #link https://review.openstack.org/#/c/563266/ | 16:26 |
| *** kumarmn_ has quit IRC | 16:26 | |
| hrybacki | tl;dr -- let's discuss the merits of making migrations idempotent as a requirement/recommendation/requirement under sitautions x, y, but not z | 16:27 |
| cmurphy | this would help with that postgres deadlock situation | 16:27 |
| hrybacki | cmurphy: I'm not following that | 16:27 |
| cmurphy | let me pull up the bug | 16:27 |
| hrybacki | ty! | 16:27 |
| hrybacki | kmalloc: made the good point that making a migration idempotent adds a level of complexity | 16:28 |
| ayoung | I'm against | 16:28 |
| kmalloc | a lot of complexity | 16:28 |
| cmurphy | https://bugs.launchpad.net/keystone/+bug/1755906 | 16:28 |
| openstack | Launchpad bug 1755906 in OpenStack Identity (keystone) "Occasional deadlock during db_sync --contract during Newton to Pike live upgrade" [Medium,Incomplete] | 16:28 |
| lbragstad | cmurphy: https://bugs.launchpad.net/keystone/+bug/1755906 ? | 16:28 |
| lbragstad | oh - yep, beat me to it :) | 16:28 |
| ayoung | Its too much work, and the migration mechanism is already designed to be Idempotent | 16:28 |
| kmalloc | ayoung: this is more for *all* migrations (going forward) | 16:28 |
| *** rossella_s has quit IRC | 16:28 | |
| kmalloc | and my guidance/recommendation is we dont | 16:29 |
| ayoung | instead, we need a better process for doing backports | 16:29 |
| kmalloc | but i am willing to humor discussing and open to changing my opinion | 16:29 |
| ayoung | I think the problem is something like this: | 16:29 |
| kmalloc | ayoung: this came up because a vendor backported | 16:29 |
| kmalloc | not upstream backported | 16:29 |
| ayoung | kmalloc, I know. I'm the vendor | 16:29 |
| kmalloc | so am i :P | 16:29 |
| *** rossella_s has joined #openstack-meeting-alt | 16:29 | |
| kmalloc | but... | 16:29 |
| kmalloc | that said | 16:29 |
| lbragstad | how complicated is doing a backport with alembic? | 16:29 |
| ayoung | we're the vendor, as is hrybacki | 16:29 |
| hrybacki | and vendors will backport. So we should think about that :) | 16:30 |
| ayoung | so...say a feature goes into R | 16:30 |
| ayoung | and we decide we want it in Q | 16:30 |
| kmalloc | i am against feature backports. full stop | 16:30 |
| ayoung | migration fills a hole in a prealloc | 16:30 |
| kmalloc | i'd argue for microversions if it stopped it | 16:30 |
| gagehugo | kmalloc heh | 16:30 |
| ayoung | but...since the feature is not in Q yet, there is no way to make the migration idempotent | 16:30 |
| kmalloc | backporting features is *bad* and should not be encourages | 16:30 |
| kmalloc | encouraged* | 16:30 |
| ayoung | so...if the vender needs that kind of idempotency, the flow should be: | 16:30 |
| ayoung | 1. submit the backport to stable. Does not need to be accepted, but should be submitted | 16:31 |
| ayoung | as a req to 1 ever getting merged... | 16:31 |
| kmalloc | lbragstad: alembic is easier, since you can just place a migration at any point (iirc) and can verify it was added/applied side-channel if it hasn't | 16:31 |
| kmalloc | but... that is a limited understanding | 16:31 |
| ayoung | 2. Submit a change to the original migration to make it idempotent | 16:31 |
| *** marios has quit IRC | 16:31 | |
| *** jchhatbar has quit IRC | 16:32 | |
| kmalloc | ayoung: i'd -2 upstream changes like that in general unless we're striving to make all migrations idempotent | 16:32 |
| ayoung | we can be lenient on the idempotent patches. They should have 0 impact | 16:32 |
| ayoung | kmalloc, what? Are you looking to make more work? | 16:32 |
| kmalloc | because migration changes could change resulting schema | 16:32 |
| ayoung | So...no | 16:32 |
| kmalloc | so. no. not more work | 16:32 |
| ayoung | the only thing we accept is a commit to an existing migration that confirms the change has already been made, and then skips | 16:32 |
| kmalloc | just avoiding accidental schema changes that end up with divergent schemas across deployments | 16:33 |
| ayoung | say we add a column in migration 101 | 16:33 |
| kmalloc | again, migrations are so touchy as is. | 16:33 |
| ayoung | the only thing we would then accept is a change to migration 101 that checks for the columnt and, if it exists, no-ops | 16:33 |
| kmalloc | changing them in the past could really really break things for some folks | 16:33 |
| ayoung | vendor submits both patches | 16:33 |
| kmalloc | even if it looks like all it does is "check" | 16:33 |
| kmalloc | my opinion: vendor carries downstream. | 16:33 |
| ayoung | there is 0 impact on the postive thread | 16:33 |
| ayoung | kmalloc, this is IFF they submit to stable | 16:34 |
| kmalloc | in the case of the bug brought up... it could have been 100% solved with a different index name | 16:34 |
| kmalloc | i don't trust them to submit to stable and master and follow through | 16:34 |
| ayoung | and that is also why it should be submitted to stable | 16:34 |
| ayoung | to get that kind of code review | 16:34 |
| kmalloc | it has to land in multiple places. | 16:34 |
| kmalloc | this feels like a really bad idea to support | 16:34 |
| hrybacki | I'm genuinely curious what the rest of the cores feel regarding vendors carrying downstream backports for migrations vs planning for that to happen in upstream | 16:34 |
| ayoung | so...we may end up NEVER accepting these, but we should document the process. | 16:35 |
| * kmalloc quiets up and lets other speak | 16:35 | |
| * ayoung too | 16:35 | |
| * hrybacki leans on others' experience and knowledge | 16:35 | |
| ayoung | and everyone looks at cmurphy! | 16:35 |
| * kmalloc stares intently at lbragstad too. | 16:36 | |
| * hrybacki rubs gagehugo's shoulders encouragingly | 16:36 | |
| lbragstad | umm | 16:36 |
| cmurphy | we mostly don't carry patches for any major features | 16:36 |
| cmurphy | we don't have any for migrations that i'm aware of | 16:36 |
| gagehugo | I don't think we carry backported features across releases | 16:36 |
| ayoung | we have, in the past, accepted changes to old migrations, when an issue has been identified that required it. | 16:37 |
| lbragstad | i'm unfamiliar with how we're managing that relationship internally | 16:37 |
| hrybacki | curious | 16:37 |
| kmalloc | ayoung: the issues were not for "we backported this" it has always been "this migration fails in this way specifically" | 16:37 |
| lbragstad | i will say that after looking into the osp issue, i see that the migration process is pretty tricky to get right | 16:37 |
| kmalloc | and those were reproducable bugs... some around things like MySQL's DDL was finacky | 16:38 |
| lbragstad | if you don't propose the original patch properly, you're kinda screwed | 16:38 |
| kmalloc | my view is pretty strongly: we don't support backporting migrations outside of legit bugs for upstream fixes. | 16:38 |
| kmalloc | if you're backporting a migration --- be aware it will be run again in upgrade. | 16:39 |
| kmalloc | and take proper steps to ensure (e.g. use a different index name) you're not causing failures. | 16:39 |
| ayoung | kmalloc, so this would be the process we would need iff we backported a fix to the stable branch | 16:39 |
| hrybacki | ^^ that scenario has to arise, right? | 16:40 |
| ayoung | I say we document it, and ensure that it is labeld "IN CASE OF FIRE BREAK GLASS" or sommat | 16:40 |
| kmalloc | we already have a way to backport migrations | 16:40 |
| kmalloc | make a migration in master (idempotent) and in a placeholder, backport placeholder | 16:40 |
| *** matrohon has quit IRC | 16:40 | |
| ayoung | kmalloc, but that will break if the migration in master is not idempotent. So Add that step in. | 16:41 |
| kmalloc | no it wont | 16:41 |
| hrybacki | so when do we decide to backport to stable or not? The bug raising this question goes way back -- at least to Liberty | 16:41 |
| kmalloc | we never ever ever ever backport a landed migration. | 16:41 |
| kmalloc | if we're doing a backport it's a BUG fix specific migration | 16:41 |
| kmalloc | and that is written to be idempotent to begin with | 16:41 |
| lbragstad | that's the step we missed with migration 22 | 16:42 |
| kmalloc | and we then land it in the placeholders (also idempotent) and then we backport just the placeholder | 16:42 |
| hrybacki | right, that iddn't happen this time -- my point | 16:42 |
| kmalloc | we didn't backport it. | 16:42 |
| ayoung | say we drop an index in master. then we realize we need it in stable. As part of that fix, we realize that id we then go from stable to master (or later stable) we are going to break. So, yeah, accepting the "check before drop" as a modification to the master migration is the only thing we can do. | 16:42 |
| kmalloc | ayoung: lets table that, you are talking about a different scenario that is not applicable here | 16:42 |
| ayoung | ++ | 16:43 |
| kmalloc | ayoung: i'm not arguing about having a mechanism to support smart migration backporting (and we have that in place -- we can better document it) | 16:43 |
| kmalloc | :) | 16:43 |
| hrybacki | right so in summary: We only 'backport' migrations when it's related to a bug that affects versions older than master? | 16:43 |
| kmalloc | the summary is: if we have a bug that requires fixing a migration there are 2 scenarios | 16:44 |
| kmalloc | 1) Bug affects everyone and we create a new migration and placeholders for bakcporting | 16:44 |
| ayoung | hrybacki, I think that, when in doubt, submit your patch to stable. And we can make the discussion public. Might not get accepted, but at least you will know why. | 16:44 |
| kmalloc | 2) the bug is in a specific migration and in some scenarios that migration is failing, but strictly in an upgrade (not a backported thing done out of tree). This is fixed in master and backported directly | 16:45 |
| kmalloc | 2 is a very very rare case | 16:45 |
| kmalloc | we have had maybe 1 of those | 16:45 |
| kmalloc | or 2 | 16:45 |
| lbragstad | ~14 minutes remaining | 16:46 |
| kmalloc | case #1 is not common, but an example is: "we need to make sure indexes are consistently named" | 16:46 |
| kmalloc | we write a migration to fix it, and placeholder backports to fix it in older releases so we're not randomly breaking due to bad assumptions | 16:46 |
| *** armaan has joined #openstack-meeting-alt | 16:46 | |
| kmalloc | everything in case #1 is idempotent | 16:46 |
| ayoung | Anything else on the Agenda? | 16:47 |
| *** kumarmn has joined #openstack-meeting-alt | 16:47 | |
| hrybacki | ack. I think I have a good understanding. (lbragstad please push us forward if there is another item on the agenda) | 16:47 |
| lbragstad | nope - that was our last topic | 16:47 |
| lbragstad | i didn't mean to cut the discussion short | 16:47 |
| hrybacki | ack | 16:48 |
| lbragstad | just keeping an eye on time is all | 16:48 |
| ayoung | OK...topic I want people to think about. We'll discuss this at the summit but... | 16:48 |
| lbragstad | #topic open discussion | 16:48 |
| *** openstack changes topic to "open discussion (Meeting topic: keystone)" | 16:48 | |
| *** salv-orl_ has quit IRC | 16:48 | |
| lbragstad | ayoung: the floor is yours | 16:48 |
| ayoung | Deleting a project today leaves all objects in remote services orphaned. What should we do about this? | 16:48 |
| ayoung | And...I think that the implied task in that is "how would a user figure out everything that they have labeled under a project to delete it" | 16:49 |
| ayoung | and also | 16:49 |
| ayoung | "how would I lock a project so I can delete stuff without letting someone else still create stuff" | 16:49 |
| ayoung | I don't think there are easy answers | 16:49 |
| ayoung | and I don't think that we can just punt on it, either, say it is not a Keystone problem | 16:50 |
| ayoung | it is not exclusively a Keystone problem, obviously | 16:50 |
| ayoung | but we should be working cross-project to come up with a range of solutions | 16:50 |
| lbragstad | i was under the impression that notifications were supposed to help with this | 16:50 |
| kmalloc | iirc, i chatted with nova and nova had a mechanism that would already allow this | 16:50 |
| kmalloc | direct API calls | 16:50 |
| kmalloc | notifications could help, but that is possibly wonky | 16:51 |
| kmalloc | since notifications could be missed. | 16:51 |
| ayoung | Issue with notifications is that the services currently have no listeners. | 16:51 |
| ayoung | So, one solution would be to build a "cleanup service" that got the project delete notification and had sufficient capabilities to delete...everything? | 16:51 |
| *** kumarmn has quit IRC | 16:52 | |
| wxy| | ayoung: ++ that's what we do in downstream. | 16:52 |
| lbragstad | so - that would still be contingent upon the notification getting received by something | 16:52 |
| ayoung | lbragstad, right,although you could also kick it off manually. | 16:53 |
| lbragstad | i suppose | 16:53 |
| lbragstad | i assume that service would need a user? | 16:53 |
| *** jessegler has quit IRC | 16:53 | |
| *** trinaths has quit IRC | 16:54 | |
| ayoung | lbragstad, I think it would be a service level role to perform, or we would have to "fake out" a project token for the deleted project | 16:54 |
| lbragstad | would deployments looking to achieve decentralized authorization from NISTs RBAC model be hesitant to adopting that kind of solution? | 16:54 |
| lbragstad | since a lot of authorization would bubble up to that role? | 16:54 |
| lbragstad | or service user? | 16:55 |
| ayoung | It is a lot of power. | 16:55 |
| lbragstad | yeah... but it's necessary if you go across the API boundaries, isn't it? | 16:55 |
| ayoung | I'd rather have it done with a soft-delete of the project, followed by cleanup, followed by a hard delete | 16:55 |
| *** sambetts is now known as sambetts|afk | 16:55 | |
| kmalloc | ayoung: hm. | 16:56 |
| lbragstad | wouldn't that require the same amount of authorization? | 16:56 |
| ayoung | lbragstad, I think "delete only tokens" might make sense once we have finer grained roles. | 16:56 |
| kmalloc | ayoung: i could probably spin up some Resource-Option fu for projects that would support "soft" delete | 16:56 |
| kmalloc | in a couple hours | 16:56 |
| kmalloc | (different semantically from "delete" or "disable") | 16:57 |
| ayoung | it could be the original role assignment that is used to determine who can delete | 16:57 |
| lbragstad | isn't disable == soft delete? | 16:57 |
| kmalloc | lbragstad: sortof. | 16:57 |
| kmalloc | not really. | 16:57 |
| *** tosky has joined #openstack-meeting-alt | 16:57 | |
| lbragstad | and we already implement that | 16:57 |
| *** Leo_m has joined #openstack-meeting-alt | 16:57 | |
| ayoung | we could say that disable is a pre-req, tho | 16:57 |
| kmalloc | it is somewhere between disable and soft-delete | 16:57 |
| ayoung | disable means you can get delete only tokens... | 16:57 |
| * ayoung waves hands | 16:58 | |
| kmalloc | because we can't change disable to waht ayoung says | 16:58 |
| kmalloc | since it changes api behavior | 16:58 |
| kmalloc | also, disable is sortof meant to be reversible | 16:58 |
| ayoung | It also leaves the question of "what do I delete where" up in the air | 16:58 |
| *** felipemonteiro__ has joined #openstack-meeting-alt | 16:59 | |
| ayoung | I mean, you could start by walking the service catalog, and say "ok, Nova...whatcha got? Delete that" | 16:59 |
| kmalloc | ayoung: i really hope we can simplify it long term to "nova - cleanup resources for project X" | 16:59 |
| ayoung | one thing to avoid is the race condition | 16:59 |
| kmalloc | not "what do you have, delete 1, delete 2, etc" | 16:59 |
| kmalloc | ;) | 16:59 |
| ayoung | "I keep trying to delte things, but kmalloc keeps making new VMs on me" | 16:59 |
| lbragstad | and with that we're out of time, | 17:00 |
| ayoung | kmalloc, so...yeah, that is exactly what I think we should work towards | 17:00 |
| hrybacki | Reminder that we are conducting our M1 retrospective. Call-in: https://bluejeans.com/8559013623 Trello Board: https://trello.com/b/PiJecAs4/keystone-rocky-m1-retrospective -- I have to break. Let's plan to join in 10? Please add cards in respective columns as you may | 17:00 |
| lbragstad | we can continue this in office hours after the retro though | 17:00 |
| lbragstad | thanks for coming | 17:00 |
| hrybacki | o/ | 17:00 |
| lbragstad | #endmeeting | 17:00 |
| *** openstack changes topic to "OpenStack Meetings || https://wiki.openstack.org/wiki/Meetings/" | 17:00 | |
| openstack | Meeting ended Tue Apr 24 17:00:45 2018 UTC. Information about MeetBot at http://wiki.debian.org/MeetBot . (v 0.1.4) | 17:00 |
| openstack | Minutes: http://eavesdrop.openstack.org/meetings/keystone/2018/keystone.2018-04-24-16.00.html | 17:00 |
| openstack | Minutes (text): http://eavesdrop.openstack.org/meetings/keystone/2018/keystone.2018-04-24-16.00.txt | 17:00 |
| openstack | Log: http://eavesdrop.openstack.org/meetings/keystone/2018/keystone.2018-04-24-16.00.log.html | 17:00 |
| *** yamamoto has joined #openstack-meeting-alt | 17:00 | |
| *** felipemonteiro__ has quit IRC | 17:01 | |
| *** derekh has quit IRC | 17:01 | |
| *** felipemonteiro__ has joined #openstack-meeting-alt | 17:01 | |
| *** felipemonteiro_ has quit IRC | 17:02 | |
| *** felipemonteiro has joined #openstack-meeting-alt | 17:03 | |
| *** iyamahat has joined #openstack-meeting-alt | 17:03 | |
| *** edmondsw has quit IRC | 17:04 | |
| *** panda|ruck is now known as panda|ruck|off | 17:04 | |
| *** yamamoto has quit IRC | 17:06 | |
| *** felipemonteiro__ has quit IRC | 17:06 | |
| *** rossella_s has quit IRC | 17:06 | |
| *** dizquierdo has joined #openstack-meeting-alt | 17:08 | |
| *** jtomasek has quit IRC | 17:10 | |
| *** rossella_s has joined #openstack-meeting-alt | 17:11 | |
| *** dsariel has quit IRC | 17:13 | |
| *** felipemonteiro_ has joined #openstack-meeting-alt | 17:17 | |
| *** felipemonteiro has quit IRC | 17:20 | |
| *** namnh has quit IRC | 17:21 | |
| *** namnh has joined #openstack-meeting-alt | 17:25 | |
| *** dkehn_ has quit IRC | 17:27 | |
| *** kumarmn has joined #openstack-meeting-alt | 17:28 | |
| *** namnh has quit IRC | 17:30 | |
| *** rmart04 has joined #openstack-meeting-alt | 17:31 | |
| *** namnh has joined #openstack-meeting-alt | 17:31 | |
| *** Leo_m_ has joined #openstack-meeting-alt | 17:33 | |
| *** kumarmn has quit IRC | 17:33 | |
| *** armaan has quit IRC | 17:33 | |
| *** lpetrut_ has joined #openstack-meeting-alt | 17:34 | |
| *** Leo_m has quit IRC | 17:36 | |
| *** Leo_m_ has quit IRC | 17:37 | |
| *** lpetrut has quit IRC | 17:38 | |
| *** kumarmn has joined #openstack-meeting-alt | 17:39 | |
| *** wxy| has quit IRC | 17:40 | |
| *** gagehugo has left #openstack-meeting-alt | 17:40 | |
| *** wxy| has joined #openstack-meeting-alt | 17:40 | |
| *** jaypipes is now known as jaypipes-lunch | 17:43 | |
| *** namnh has quit IRC | 17:44 | |
| *** rmart04 has quit IRC | 17:45 | |
| *** namnh has joined #openstack-meeting-alt | 17:45 | |
| *** rmart04 has joined #openstack-meeting-alt | 17:46 | |
| *** namnh has quit IRC | 17:55 | |
| *** wxy| has quit IRC | 17:55 | |
| *** bhavik1 has quit IRC | 17:55 | |
| *** vgreen has quit IRC | 17:55 | |
| *** edmondsw has joined #openstack-meeting-alt | 17:56 | |
| *** kumarmn has quit IRC | 17:59 | |
| *** namnh has joined #openstack-meeting-alt | 18:00 | |
| *** dizquierdo has quit IRC | 18:02 | |
| *** yamamoto has joined #openstack-meeting-alt | 18:03 | |
| *** edmondsw_ has joined #openstack-meeting-alt | 18:03 | |
| *** edmonds__ has joined #openstack-meeting-alt | 18:05 | |
| *** edmonds__ is now known as edmondsw__ | 18:06 | |
| *** edmondsw__ is now known as edmondsw___ | 18:06 | |
| *** edmondsw has quit IRC | 18:06 | |
| *** edmondsw___ is now known as edmondsw | 18:06 | |
| *** yamamoto has quit IRC | 18:08 | |
| *** edmondsw_ has quit IRC | 18:09 | |
| *** namnh has quit IRC | 18:10 | |
| *** cloudrancher has joined #openstack-meeting-alt | 18:16 | |
| *** lpetrut_ has quit IRC | 18:17 | |
| *** kumarmn has joined #openstack-meeting-alt | 18:20 | |
| *** jaypipes-lunch is now known as jaypipes | 18:23 | |
| *** kumarmn has quit IRC | 18:24 | |
| *** ianychoi has quit IRC | 18:26 | |
| *** iyamahat has quit IRC | 18:26 | |
| *** ianychoi has joined #openstack-meeting-alt | 18:26 | |
| *** gyee has quit IRC | 18:28 | |
| *** cloudrancher has quit IRC | 18:31 | |
| *** felipemonteiro__ has joined #openstack-meeting-alt | 18:39 | |
| *** felipemonteiro_ has quit IRC | 18:39 | |
| *** rmart04 has quit IRC | 18:53 | |
| *** fnaval has joined #openstack-meeting-alt | 18:54 | |
| *** sridharg has quit IRC | 18:59 | |
| *** kumarmn has joined #openstack-meeting-alt | 19:01 | |
| *** dims has quit IRC | 19:02 | |
| *** yamahata has joined #openstack-meeting-alt | 19:04 | |
| *** yamamoto has joined #openstack-meeting-alt | 19:04 | |
| *** lpetrut_ has joined #openstack-meeting-alt | 19:04 | |
| *** iyamahat has joined #openstack-meeting-alt | 19:05 | |
| *** dims has joined #openstack-meeting-alt | 19:09 | |
| *** yamamoto has quit IRC | 19:10 | |
| *** radeks has quit IRC | 19:14 | |
| *** matrohon has joined #openstack-meeting-alt | 19:14 | |
| *** salv-orlando has joined #openstack-meeting-alt | 19:16 | |
| *** Leo_m has joined #openstack-meeting-alt | 19:20 | |
| *** Leo_m has quit IRC | 19:20 | |
| *** Leo_m has joined #openstack-meeting-alt | 19:20 | |
| *** Leo_m has quit IRC | 19:25 | |
| *** tesseract has quit IRC | 19:26 | |
| *** Leo_m has joined #openstack-meeting-alt | 19:32 | |
| *** ianychoi has quit IRC | 19:39 | |
| *** ianychoi has joined #openstack-meeting-alt | 19:41 | |
| *** matrohon has quit IRC | 19:58 | |
| *** matrohon has joined #openstack-meeting-alt | 19:59 | |
| *** kumarmn has quit IRC | 20:01 | |
| *** kumarmn has joined #openstack-meeting-alt | 20:02 | |
| *** rasca has quit IRC | 20:04 | |
| *** weshay has quit IRC | 20:05 | |
| *** ianychoi has quit IRC | 20:05 | |
| *** yamamoto has joined #openstack-meeting-alt | 20:06 | |
| *** lpetrut_ has quit IRC | 20:07 | |
| *** weshay has joined #openstack-meeting-alt | 20:07 | |
| *** rasca has joined #openstack-meeting-alt | 20:07 | |
| *** spilla has left #openstack-meeting-alt | 20:07 | |
| *** ianychoi has joined #openstack-meeting-alt | 20:09 | |
| *** yamamoto has quit IRC | 20:11 | |
| *** dklyle has quit IRC | 20:18 | |
| *** cloudrancher has joined #openstack-meeting-alt | 20:19 | |
| *** AlanClark has quit IRC | 20:24 | |
| *** kumarmn has quit IRC | 20:24 | |
| *** slaweq has quit IRC | 20:25 | |
| *** slaweq has joined #openstack-meeting-alt | 20:26 | |
| *** iyamahat has quit IRC | 20:28 | |
| *** cloudrancher has quit IRC | 20:29 | |
| *** yamahata has quit IRC | 20:29 | |
| *** kumarmn has joined #openstack-meeting-alt | 20:29 | |
| *** priteau_ has joined #openstack-meeting-alt | 20:31 | |
| *** spzala has quit IRC | 20:32 | |
| *** spzala has joined #openstack-meeting-alt | 20:32 | |
| *** raildo has quit IRC | 20:34 | |
| *** raildo has joined #openstack-meeting-alt | 20:34 | |
| *** jcoufal_ has joined #openstack-meeting-alt | 20:34 | |
| *** priteau has quit IRC | 20:34 | |
| *** kumarmn has quit IRC | 20:34 | |
| *** spzala has quit IRC | 20:34 | |
| *** stevebaker has quit IRC | 20:37 | |
| *** jcoufal has quit IRC | 20:37 | |
| *** gouthamr has quit IRC | 20:38 | |
| *** alexchadin has joined #openstack-meeting-alt | 20:38 | |
| *** alexchadin has quit IRC | 20:38 | |
| *** ianychoi has quit IRC | 20:39 | |
| *** ianychoi has joined #openstack-meeting-alt | 20:40 | |
| *** stevebaker has joined #openstack-meeting-alt | 20:42 | |
| *** gouthamr has joined #openstack-meeting-alt | 20:43 | |
| *** rmascena has joined #openstack-meeting-alt | 20:44 | |
| *** raildo has quit IRC | 20:46 | |
| *** rmascena has quit IRC | 20:50 | |
| *** iyamahat has joined #openstack-meeting-alt | 20:50 | |
| *** iyamahat_ has joined #openstack-meeting-alt | 20:50 | |
| *** iyamahat has quit IRC | 20:55 | |
| *** dklyle has joined #openstack-meeting-alt | 20:56 | |
| *** kumarmn has joined #openstack-meeting-alt | 20:58 | |
| *** cloudrancher has joined #openstack-meeting-alt | 21:05 | |
| *** cloudrancher has quit IRC | 21:05 | |
| *** yamamoto has joined #openstack-meeting-alt | 21:08 | |
| *** yamahata has joined #openstack-meeting-alt | 21:09 | |
| *** edmondsw_ has joined #openstack-meeting-alt | 21:11 | |
| *** edmondsw has quit IRC | 21:13 | |
| *** yamamoto has quit IRC | 21:14 | |
| *** dklyle has quit IRC | 21:14 | |
| *** slaweq has quit IRC | 21:14 | |
| *** slaweq has joined #openstack-meeting-alt | 21:15 | |
| *** slaweq has quit IRC | 21:20 | |
| *** matrohon has quit IRC | 21:21 | |
| *** fnaval has quit IRC | 21:21 | |
| *** ianychoi has quit IRC | 21:24 | |
| *** ianychoi has joined #openstack-meeting-alt | 21:25 | |
| *** jcoufal_ has quit IRC | 21:27 | |
| *** fnaval has joined #openstack-meeting-alt | 21:32 | |
| *** priteau_ has quit IRC | 21:32 | |
| *** ianychoi has quit IRC | 21:35 | |
| *** ianychoi has joined #openstack-meeting-alt | 21:36 | |
| *** namnh has joined #openstack-meeting-alt | 21:37 | |
| *** jbadiapa has joined #openstack-meeting-alt | 21:38 | |
| *** cloudrancher has joined #openstack-meeting-alt | 21:41 | |
| *** dklyle has joined #openstack-meeting-alt | 21:44 | |
| *** cloudrancher has quit IRC | 21:46 | |
| *** edmondsw_ has quit IRC | 21:55 | |
| *** ianychoi has quit IRC | 22:02 | |
| *** ianychoi has joined #openstack-meeting-alt | 22:04 | |
| *** kumarmn has quit IRC | 22:04 | |
| *** edmondsw has joined #openstack-meeting-alt | 22:05 | |
| *** edmondsw has quit IRC | 22:05 | |
| *** namnh has quit IRC | 22:07 | |
| *** yamamoto has joined #openstack-meeting-alt | 22:09 | |
| *** yamamoto has quit IRC | 22:14 | |
| *** dave-mccowan has quit IRC | 22:15 | |
| *** gyee has joined #openstack-meeting-alt | 22:23 | |
| *** dtrainor has quit IRC | 22:41 | |
| *** efried has quit IRC | 22:42 | |
| *** fnaval has quit IRC | 22:45 | |
| *** dizquierdo has joined #openstack-meeting-alt | 22:52 | |
| *** fnaval has joined #openstack-meeting-alt | 22:53 | |
| *** fnaval has quit IRC | 22:54 | |
| *** fnaval has joined #openstack-meeting-alt | 22:54 | |
| *** dustins has quit IRC | 22:58 | |
| *** dizquierdo has quit IRC | 23:09 | |
| *** yamamoto has joined #openstack-meeting-alt | 23:10 | |
| *** cloudrancher has joined #openstack-meeting-alt | 23:10 | |
| *** cloudrancher has quit IRC | 23:15 | |
| *** felipemonteiro__ has quit IRC | 23:16 | |
| *** yamamoto has quit IRC | 23:16 | |
| *** cloudrancher has joined #openstack-meeting-alt | 23:18 | |
| *** cloudrancher has quit IRC | 23:22 | |
| *** cloudrancher has joined #openstack-meeting-alt | 23:23 | |
| *** dizquierdo has joined #openstack-meeting-alt | 23:25 | |
| *** chyka_ has quit IRC | 23:29 | |
| *** cloudrancher has quit IRC | 23:32 | |
| *** hongbin_ has quit IRC | 23:33 | |
| *** cloudrancher has joined #openstack-meeting-alt | 23:36 | |
| *** tpsilva has quit IRC | 23:36 | |
| *** dizquierdo has quit IRC | 23:39 | |
| *** cloudrancher has quit IRC | 23:55 | |
| *** tosky has quit IRC | 23:58 | |
| *** yamamoto has joined #openstack-meeting-alt | 23:58 | |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!