| *** yingjun has joined #openstack-searchlight | 00:16 | |
| openstackgerrit | Travis Tripp proposed openstack/searchlight: Enable setting listener thread pool size https://review.openstack.org/317138 | 00:46 |
|---|---|---|
| *** bpokorny_ has joined #openstack-searchlight | 01:03 | |
| *** bpokorny has quit IRC | 01:07 | |
| *** bpokorny_ has quit IRC | 01:08 | |
| *** itisha has quit IRC | 01:29 | |
| *** shu-mutou-AFK is now known as shu-mutou | 01:40 | |
| openstackgerrit | Lakshmi N Sampath proposed openstack/searchlight: Resource types for reference and parent id https://review.openstack.org/317167 | 02:27 |
| *** lakshmiS has quit IRC | 02:31 | |
| *** bpokorny has joined #openstack-searchlight | 04:11 | |
| *** lakshmiS has joined #openstack-searchlight | 04:12 | |
| *** bpokorny has quit IRC | 04:20 | |
| openstackgerrit | Lakshmi N Sampath proposed openstack/searchlight: Resource types for reference and parent id https://review.openstack.org/317167 | 04:37 |
| *** lakshmiS has quit IRC | 04:45 | |
| *** GB21 has joined #openstack-searchlight | 04:56 | |
| *** GB21 has quit IRC | 05:04 | |
| *** GB21 has joined #openstack-searchlight | 05:09 | |
| *** GB21 has quit IRC | 05:31 | |
| *** yingjun has quit IRC | 09:39 | |
| *** shu-mutou is now known as shu-mutou-AFK | 09:48 | |
| *** TravT has quit IRC | 11:06 | |
| *** TravT has joined #openstack-searchlight | 11:07 | |
| *** TravT_ has joined #openstack-searchlight | 11:49 | |
| *** TravT has quit IRC | 11:51 | |
| *** yingjun has joined #openstack-searchlight | 13:01 | |
| *** openstackgerrit has quit IRC | 13:32 | |
| *** openstackgerrit has joined #openstack-searchlight | 13:34 | |
| *** sigmavirus24_awa is now known as sigmavirus24 | 13:54 | |
| *** jasongalyon has joined #openstack-searchlight | 13:56 | |
| openstackgerrit | OpenStack Proposal Bot proposed openstack/searchlight: Updated from global requirements https://review.openstack.org/316957 | 14:10 |
| openstackgerrit | Steve McLellan proposed openstack/searchlight: Add totals to facets https://review.openstack.org/308582 | 15:24 |
| *** TravT_ is now known as TravT | 15:25 | |
| *** tyr_ has joined #openstack-searchlight | 15:29 | |
| *** tyr__ has joined #openstack-searchlight | 15:30 | |
| *** tyr_ has quit IRC | 15:30 | |
| openstackgerrit | Travis Tripp proposed openstack/searchlight: Enable setting listener thread pool size https://review.openstack.org/317138 | 15:33 |
| *** yingjun has quit IRC | 15:45 | |
| *** yingjun has joined #openstack-searchlight | 15:45 | |
| *** lakshmiS has joined #openstack-searchlight | 15:46 | |
| *** yingjun has quit IRC | 15:50 | |
| *** bpokorny has joined #openstack-searchlight | 15:52 | |
| TravT | sjmc7: if you have a second, can you weigh in on this patch? | 15:55 |
| TravT | https://review.openstack.org/#/c/317138/ | 15:56 |
| TravT | need to decide whether to keep it or abandon it | 15:56 |
| tyr__ | morning all! I'm working on untangling searchlight-ui's use of events, and the horrible event system in magic search. It makes searchlight-ui and magic search MUCH easier to work with and use in creative ways. | 15:56 |
| tyr__ | The works will come in as a series of about 5 patches. I could use some help clearing the first easy ones. | 15:56 |
| sjmc7 | yeah, will do - although that messiness made me realize that the nova problem was more complicated than i thought :) | 15:56 |
| tyr__ | First easy one: https://review.openstack.org/317073 | 15:57 |
| tyr__ | Second easy one: https://review.openstack.org/317121 | 15:57 |
| TravT | tyr__: i was just about to look at the easy ones. thanks! | 15:58 |
| sjmc7 | TravT: the executor_thread_pool_size option from oslo will end up under DEFAULT? | 16:01 |
| sjmc7 | yeah, looks like it | 16:03 |
| TravT | yes | 16:10 |
| TravT | i'm not sure if that also effects API or not | 16:10 |
| TravT | sjmc7: it looks like we won't be getting any help from nova versioned notifications providing any more data. | 16:18 |
| TravT | https://review.openstack.org/#/c/286675/16 | 16:18 |
| TravT | Gibi seems to be actively advocating against putting more data in them | 16:19 |
| TravT | so, they will continue to be problematic. | 16:19 |
| openstackgerrit | Lakshmi N Sampath proposed openstack/searchlight: Resource types for reference and parent id https://review.openstack.org/317167 | 16:21 |
| sjmc7 | ok. so there’s not really any benefit in them | 16:24 |
| TravT | as far as i can tell, there doesn't seem to be much | 16:27 |
| TravT | only that you will be able to tell when they break you | 16:27 |
| TravT | using version | 16:28 |
| TravT | and the schema will give you the mapping | 16:28 |
| TravT | for string or what not | 16:28 |
| TravT | i'm tempted to go to the meeting and tell them that i've lost sight of what value the versioned notifications bring | 16:28 |
| sjmc7 | what can break if they’re not adding stuff? | 16:28 |
| sjmc7 | unless they’re going to remove things | 16:29 |
| sjmc7 | in which case they should just do it | 16:29 |
| TravT | in theory the notifications follow semver | 16:29 |
| TravT | i need to re-read their use cases. | 16:29 |
| TravT | this is all part of their rolling upgrade work somehow | 16:30 |
| sjmc7 | on a happier note, i took a look at messaging pools earlier and i think it can simplify deployment | 16:33 |
| *** tyr_ has joined #openstack-searchlight | 16:41 | |
| *** tyr__ has quit IRC | 16:43 | |
| *** tyr_ has quit IRC | 16:54 | |
| *** tyr_ has joined #openstack-searchlight | 16:55 | |
| lakshmiS | TravT: hey | 17:07 |
| TravT | hey lakshmiS | 17:07 |
| lakshmiS | for https://bugs.launchpad.net/searchlight/+bug/1569485 | 17:08 |
| openstack | Launchpad bug 1569485 in OpenStack Search (Searchlight) "Image type should be a facet - Can't select snapshots vs images" [High,In progress] - Assigned to Lakshmi N Sampath (lakshmi-sampath) | 17:08 |
| lakshmiS | Image based on snapshot has these additional fields | 17:08 |
| lakshmiS | "image_state": "available", | 17:08 |
| lakshmiS | "user_id": "82b5649a53bf4794bc464bbf3db93a57", | 17:08 |
| lakshmiS | "image_type": "snapshot", | 17:08 |
| lakshmiS | "base_image_ref": "62c26a7b-fd5c-4619-869e-5fd87d69eb2f", | 17:08 |
| lakshmiS | "image_location": "snapshot", | 17:08 |
| lakshmiS | "instance_uuid": "e60db0b7-052e-4f18-9a67-88ef20166682", | 17:08 |
| lakshmiS | I will add image_type and base_image_ref(link) for mapping | 17:08 |
| lakshmiS | do you see anything else here useful? | 17:09 |
| TravT | hmmm... all of those look useful. | 17:09 |
| lakshmiS | i meant for facet listing | 17:09 |
| lakshmiS | being dynamic mapping they are indexed anyway | 17:10 |
| lakshmiS | but would UI users search based on user_id or instance_id? | 17:10 |
| lakshmiS | image_state/user_id/instance_uuid would default to null if its a non-snapshot image | 17:13 |
| TravT | hey, just a sec... i'm in the nova notification meeting and they just started talking about the comments i left on the spec | 17:13 |
| lakshmiS | k | 17:13 |
| TravT | lakshmiS: that is really interesting re: facets | 17:35 |
| TravT | really makes me wonder if we should break the mold and have OS::Glance::Snapshot | 17:35 |
| lakshmiS | hmm.. its not completely different either since we will still handle same notification event and api for both | 17:37 |
| TravT | so, openstack client has snapshot list | 17:39 |
| TravT | in addition to image list | 17:39 |
| TravT | lakshmiS: I don't think image_location is worth indexing of the above | 17:40 |
| lakshmiS | yeah i removed it | 17:40 |
| lakshmiS | also image_state might get confusing with 'status' field | 17:41 |
| lakshmiS | i think its good to differentiate image vs snapshot from user point of view but from our code perspective its same code for mapping/notfications/etc except for those 5 fields | 17:42 |
| lakshmiS | I can give it a try by extending ImageIndex | 17:43 |
| TravT | rosmaita are you around? | 17:44 |
| TravT | lakshmiS: openstack client snapshot list doesn't return anything | 17:47 |
| TravT | i wonder if that is a volume snapshot | 17:47 |
| lakshmiS | could be. let me check code | 17:47 |
| TravT | my image snapshot still comes back from image list | 17:47 |
| lakshmiS | well even if glance doesn't treat it separately, we could still have OS::Glance::Snapshot | 17:48 |
| rosmaita | TravT: hello | 17:49 |
| TravT | hey rosmaita we were just debating on the merits of keeping images and snapshots under one resource type OS::Glance::Image vs also having OS::Glance::Snapshot | 17:51 |
| TravT | the available facets differ based on which one you care about. | 17:51 |
| rosmaita | glance doesn't recognize a difference | 17:52 |
| lakshmiS | seems like those fields are just metadata added to glance image | 17:52 |
| TravT | yeah, we probably should just keep as image then. | 17:52 |
| rosmaita | i think so | 17:52 |
| rosmaita | some of those fileds lakshmiS quoted above look non-standard to me | 17:53 |
| lakshmiS | then we should accept null value for snapshot fields in image data and UI users will have some confusion over snapshot fields :( | 17:53 |
| rosmaita | lakshmiS: what exactly are "snapshot fields"? | 17:53 |
| lakshmiS | the list mentioned above which are part of API result only if image is based on a snapshot | 17:54 |
| rosmaita | image_type i think is defined in the compute api | 17:54 |
| rosmaita | for nova, you can have base, snapshot, or backup | 17:54 |
| rosmaita | at rackspace, we also put 'import' on imported images | 17:55 |
| lakshmiS | interesting, we were thinking of adding 'image' as default value for vanilla imagews | 17:55 |
| rosmaita | i think in most clouds all you'll see are 'base' or 'snapshot' | 17:56 |
| lakshmiS | rosmaita: do you know what is 'image_state' possible values? and i guess this is completely different from 'status' field on glance image | 17:56 |
| TravT | rosmaita: FYI, this is the bug being referenced: https://bugs.launchpad.net/searchlight/+bug/1569485 | 17:57 |
| openstack | Launchpad bug 1569485 in OpenStack Search (Searchlight) "Image type should be a facet - Can't select snapshots vs images" [High,In progress] - Assigned to Lakshmi N Sampath (lakshmi-sampath) | 17:57 |
| TravT | maybe you could add your thoughts on there? | 17:57 |
| rosmaita | ok, will do | 17:57 |
| rosmaita | base_image_ref is definitely useful | 17:58 |
| lakshmiS | yes | 17:58 |
| lakshmiS | TravT: only hit for OS::Glance::Snapshot on google is Graffit :) | 18:00 |
| TravT | haha | 18:00 |
| TravT | yeah, so let's stay with OS::Glance::Image | 18:00 |
| TravT | and just give facets. | 18:01 |
| rosmaita | TravT: added a comment to that bug | 18:14 |
| *** tyr_ has quit IRC | 18:27 | |
| *** tyr_ has joined #openstack-searchlight | 18:27 | |
| lakshmiS | thanks | 18:32 |
| *** jasongalyon has quit IRC | 19:31 | |
| *** jasongalyon has joined #openstack-searchlight | 19:31 | |
| *** bpokorny has quit IRC | 20:03 | |
| *** jasongalyon has quit IRC | 20:12 | |
| *** jasongalyon has joined #openstack-searchlight | 20:57 | |
| *** bpokorny has joined #openstack-searchlight | 21:17 | |
| openstackgerrit | Lakshmi N Sampath proposed openstack/searchlight: WIP: Image type should be a facet https://review.openstack.org/317741 | 21:28 |
| *** lakshmiS has quit IRC | 21:35 | |
| *** jasongalyon has quit IRC | 21:53 | |
| *** sigmavirus24 is now known as sigmavirus24_awa | 22:08 | |
| openstackgerrit | Tyr Johanson proposed openstack/searchlight-ui: Searchlight version of Magic Search https://review.openstack.org/317133 | 22:30 |
| openstackgerrit | Tyr Johanson proposed openstack/searchlight-ui: Searchlight version of Magic Search https://review.openstack.org/317133 | 22:59 |
| openstackgerrit | Tyr Johanson proposed openstack/searchlight-ui: Searchlight version of hz-expand-detail https://review.openstack.org/317121 | 23:00 |
| openstackgerrit | Tyr Johanson proposed openstack/searchlight-ui: WIP Integrating sl-magic-search and searchlight-ui https://review.openstack.org/317760 | 23:02 |
| *** tyr_ has quit IRC | 23:06 | |
| *** lakshmiS has joined #openstack-searchlight | 23:08 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!