| *** RickA-HP has quit IRC | 00:08 | |
| openstackgerrit | Rick Aulino proposed openstack/searchlight: To speed up Searchlight indexing and re-indexing we should use threads instead of inexing resource types sequentially. https://review.openstack.org/333649 | 01:32 |
|---|---|---|
| *** lei-zh has joined #openstack-searchlight | 02:23 | |
| *** shu-mutou has quit IRC | 02:24 | |
| openstackgerrit | Tang Chen proposed openstack/python-searchlightclient: Use osc-lib instead of openstackclient https://review.openstack.org/333661 | 02:47 |
| *** M00nr41n has joined #openstack-searchlight | 02:53 | |
| openstackgerrit | Tang Chen proposed openstack/python-searchlightclient: Use osc-lib instead of openstackclient https://review.openstack.org/333661 | 02:58 |
| *** shu-mutou has joined #openstack-searchlight | 04:03 | |
| *** M00nr41n has quit IRC | 04:03 | |
| *** GB21 has joined #openstack-searchlight | 04:15 | |
| *** david-lyle has quit IRC | 04:20 | |
| *** david-lyle has joined #openstack-searchlight | 04:20 | |
| openstackgerrit | GB21 proposed openstack/searchlight: Flavor Plugin https://review.openstack.org/315409 | 04:34 |
| openstackgerrit | GB21 proposed openstack/searchlight: Flavor Plugin https://review.openstack.org/315409 | 04:57 |
| *** M00nr41n has joined #openstack-searchlight | 05:04 | |
| *** GB21 has quit IRC | 05:05 | |
| *** GB21 has joined #openstack-searchlight | 06:04 | |
| *** GB21 has quit IRC | 06:12 | |
| *** GB21 has joined #openstack-searchlight | 06:43 | |
| *** shu-mutou has quit IRC | 07:00 | |
| *** lei-zh1 has joined #openstack-searchlight | 08:01 | |
| *** lei-zh has quit IRC | 08:04 | |
| *** GB21 has quit IRC | 08:45 | |
| *** lei-zh1 has quit IRC | 09:01 | |
| openstackgerrit | Zhenyu Zheng proposed openstack/searchlight: Handle JSON object in hypervisor serialization https://review.openstack.org/333832 | 09:12 |
| *** pcaruana has joined #openstack-searchlight | 10:09 | |
| *** M00nr41n has quit IRC | 11:20 | |
| *** ddieterly has joined #openstack-searchlight | 11:53 | |
| *** ddieterly has quit IRC | 11:59 | |
| *** ddieterly has joined #openstack-searchlight | 12:11 | |
| *** ddieterly has quit IRC | 12:26 | |
| *** ddieterly has joined #openstack-searchlight | 12:46 | |
| *** ddieterly has quit IRC | 13:02 | |
| *** ddieterly has joined #openstack-searchlight | 13:33 | |
| *** ddieterly is now known as ddieterly[away] | 13:40 | |
| *** ddieterly[away] is now known as ddieterly | 13:47 | |
| *** TravT has joined #openstack-searchlight | 14:09 | |
| *** TravT has quit IRC | 14:09 | |
| *** TravT has joined #openstack-searchlight | 14:13 | |
| *** TravT has quit IRC | 14:13 | |
| *** sjmc7 has joined #openstack-searchlight | 14:14 | |
| *** TravT has joined #openstack-searchlight | 14:14 | |
| openstackgerrit | Rick Aulino proposed openstack/searchlight: Add Neutron security group plugin https://review.openstack.org/326767 | 14:26 |
| *** RickA-HP has joined #openstack-searchlight | 14:31 | |
| RickA-HP | Folks, https://review.openstack.org/#/c/333525 (ES2.3.2 testing) was approved by the infra team and merged last night. This means Zuul is now mainly testing the patches against ES2.3.2. | 14:33 |
| RickA-HP | As a reminder we are testing: Py27/ES2.3.2, Py27/ES1.7.5 and Py34/ES2.3.2. | 14:33 |
| sjmc7 | cool, good to get that change made. do we have a bp/bug to try to upgrade the elasticsearch client version? | 14:34 |
| RickA-HP | sjmc7: Yes, I submitted it yesterday. | 14:39 |
| sjmc7 | cool. hopefully there won’t be anything that’s not back-compatible | 14:40 |
| RickA-HP | I also just kicked off a patch for https://review.openstack.org/#/c/326767 to verify the new Zuul testing. Everything passed with the new configurations. | 14:40 |
| *** ddieterly is now known as ddieterly[away] | 14:58 | |
| openstackgerrit | Steve McLellan proposed openstack/searchlight: Allow dynamic index settings https://review.openstack.org/333958 | 15:17 |
| *** ddieterly[away] is now known as ddieterly | 15:32 | |
| *** pcaruana has quit IRC | 15:38 | |
| *** itisha has quit IRC | 15:50 | |
| *** ddieterly is now known as ddieterly[away] | 15:55 | |
| *** ddieterly[away] is now known as ddieterly | 15:58 | |
| *** ddieterly has quit IRC | 16:44 | |
| *** GB21 has joined #openstack-searchlight | 17:30 | |
| GB21 | hi TravT and sjmc7 | 17:31 |
| sjmc7 | hi | 17:31 |
| TravT | hi | 17:31 |
| openstackgerrit | Merged openstack/searchlight: Updated from global requirements https://review.openstack.org/333703 | 17:40 |
| GB21 | I was thinking if I could start the heat plugin | 17:41 |
| sjmc7 | GB21: i’d started looking at that already. there are probably some bugs you could pick up? | 17:43 |
| TravT | we were just chatting the other day that we have 53 bugs outstanding and want to get some of them closed out. | 17:44 |
| GB21 | sjmc7, okay, I certainly can; also there was a blueprint to asynchronize elasticsearch | 17:44 |
| GB21 | ohh | 17:44 |
| GB21 | okay | 17:45 |
| GB21 | I will start working on bugs then | 17:45 |
| TravT | that would be really awesome. | 17:45 |
| sjmc7 | and you may well also find more bugs :) | 17:45 |
| TravT | just thinking that we can work on some of our technical debt before adding too many more plugins | 17:45 |
| GB21 | hehe, technical debt? | 17:47 |
| GB21 | ohh, cool | 17:49 |
| GB21 | I will start working on bugs then. | 17:49 |
| TravT | thanks GB21! | 17:50 |
| GB21 | no issues TravT :D | 17:50 |
| openstackgerrit | Matt Borland proposed openstack/searchlight-ui: Adding registry-based drawer (summary) views https://review.openstack.org/332154 | 17:50 |
| GB21 | well, I will be pinging sjmc7 more :P | 17:50 |
| GB21 | Also, please review my latest patch | 17:51 |
| GB21 | Thanks | 17:51 |
| sjmc7 | will do | 17:52 |
| GB21 | Thank you :D | 17:56 |
| TravT | so sjmc7 https://review.openstack.org/#/c/332242/ | 17:56 |
| sjmc7 | yes | 17:57 |
| TravT | i use the top level type as a convenience for indexing all of them quite often. | 17:57 |
| sjmc7 | ok | 17:57 |
| TravT | e.g. --type OS::Neutron::Network | 17:57 |
| sjmc7 | that’s why i added —include-children originally and everyone moaned :) | 17:57 |
| TravT | well, its because it still was harder to use... ;) | 17:58 |
| sjmc7 | it defaulted to on | 17:58 |
| TravT | i had to remember that there were children | 17:58 |
| TravT | and it seemed to still fail for me | 17:58 |
| TravT | maybe i tested it wrong | 17:58 |
| sjmc7 | ok. what behavior would you like? | 17:58 |
| TravT | more glowing praise | 17:59 |
| TravT | maybe if you can roll over or sit on command. | 17:59 |
| TravT | you know, all the standard stuff | 17:59 |
| TravT | ;) | 17:59 |
| sjmc7 | the three cases are ‘index this’, ‘index this and include any children it has’, ‘index everything’ | 17:59 |
| TravT | so, maybe i totally looked at it wrong the first time. i guess both yingjun and i were confused by it | 18:00 |
| sjmc7 | ok. i should’ve asked you to clarify the expected behavior on the bug, so that’s my bad | 18:00 |
| TravT | if you specified OS::Neutron::Net and all the children indexed by default and worked without getting that error, i'd be happy. | 18:01 |
| TravT | the option you added to exclude children would be fine | 18:01 |
| sjmc7 | ok | 18:01 |
| TravT | but i thought i still saw that version conflict bug | 18:01 |
| sjmc7 | there’s no technical reason to index ports just because networks happens to be their parent | 18:02 |
| TravT | well, it came down to knowing why you are re-indexing. | 18:02 |
| TravT | it seems like only in remapping or bug fix scenarios would we index just one thing | 18:03 |
| *** openstackgerrit has quit IRC | 18:03 | |
| TravT | i mean that we'd know to re-index just ports | 18:03 |
| TravT | or just networks | 18:03 |
| sjmc7 | ok | 18:03 |
| TravT | otherwise, i'm just wondering how an admin would know? | 18:03 |
| TravT | what do you think? | 18:03 |
| sjmc7 | that suggests not triggering a reindexing of loads of stuff just because you wanted to reindex networks | 18:03 |
| *** openstackgerrit has joined #openstack-searchlight | 18:03 | |
| TravT | yes, the exclude children for fixing a network mapping case makes sense | 18:04 |
| TravT | but the "I think there's a coherency problem" so i better reindex X seems hard to know what to do. | 18:04 |
| sjmc7 | i think i’d argue that we shouldn’t have the option at all; if you want to reindex networks and ports for instance, do them both | 18:04 |
| TravT | but the root problem I really saw was that I thought the version conflict still failed | 18:04 |
| TravT | on re-indexing with children | 18:05 |
| sjmc7 | ok, that’s obviously something would need fixing | 18:05 |
| TravT | i can try out this patch again | 18:06 |
| sjmc7 | no, i’m sure you’re right that there was a bug | 18:06 |
| TravT | if it still has failures then it still is a problem | 18:06 |
| sjmc7 | i am not disputingh that | 18:06 |
| sjmc7 | i want to make sure the behavior makes sense | 18:06 |
| TravT | otherwise, i'll think a bit about the behavior | 18:06 |
| sjmc7 | let’s decide on behavior first :) | 18:06 |
| sjmc7 | i posit that reindexing four RTs because you happened to want to reindex networks is a bit odd | 18:07 |
| TravT | i can see your argument for --include-children | 18:07 |
| TravT | i don't dispute that | 18:07 |
| TravT | just want to make sure it is easy to get children without failure | 18:07 |
| sjmc7 | with that patch as-is, there’s no option to ‘index X and all children of X’ | 18:08 |
| TravT | i'm not sure which way is more error prone | 18:08 |
| sjmc7 | so if there is a use case for that we could add it | 18:08 |
| sjmc7 | i mean, indexing ports separate from servers could lead to inconsistency too | 18:09 |
| sjmc7 | the parent/child thing is an implementation detail | 18:09 |
| TravT | yes, it is | 18:09 |
| sjmc7 | ok. well, try it out | 18:10 |
| sjmc7 | i’m gonna eat | 18:10 |
| TravT | nom nom | 18:10 |
| *** GB21 has quit IRC | 18:24 | |
| *** openstackstatus has quit IRC | 18:40 | |
| *** openstack has joined #openstack-searchlight | 18:45 | |
| *** sjmc7 has quit IRC | 18:59 | |
| *** sjmc7 has joined #openstack-searchlight | 19:16 | |
| *** RickA-HP has quit IRC | 19:24 | |
| TravT | sjmc7 left thoughts on that patch | 19:26 |
| TravT | i'm gonna go eat now | 19:27 |
| sjmc7 | okey dokey | 19:27 |
| openstackgerrit | Steve McLellan proposed openstack/searchlight: Fix --type handling for indexing parent plugins https://review.openstack.org/332242 | 20:03 |
| openstackgerrit | Matt Borland proposed openstack/searchlight-ui: Adding registry-based drawer (summary) views https://review.openstack.org/332154 | 20:47 |
| openstackgerrit | Steve McLellan proposed openstack/searchlight: Fix --type handling for indexing parent plugins https://review.openstack.org/332242 | 20:58 |
| openstackgerrit | Matt Borland proposed openstack/searchlight-ui: Adding registry-based drawer (summary) views https://review.openstack.org/332154 | 21:04 |
| openstackgerrit | Travis Tripp proposed openstack/searchlight: Reduce number of API calls on nova boot https://review.openstack.org/317100 | 21:36 |
| TravT | sjmc7: ^ just rebasing to force the zuul recheck | 21:37 |
| sjmc7 | k | 21:38 |
| openstackgerrit | Merged openstack/searchlight-ui: Adding registry-based drawer (summary) views https://review.openstack.org/332154 | 21:51 |
| sjmc7 | TravT: gonna register BPs for the additional indexing behavior | 22:04 |
| TravT | okay, thx | 22:05 |
| sjmc7 | i keep forgetting to add release notes to git commits | 22:12 |
| sjmc7 | it’s like i have a mental block against them | 22:12 |
| openstackgerrit | Steve McLellan proposed openstack/searchlight: Fix --type handling for indexing parent plugins https://review.openstack.org/332242 | 22:15 |
| sjmc7 | TravT: number_of_shards is apparently at index creation time only; i’ll have to see if we can pass settings through as part of index creation rather than separately | 22:20 |
| sjmc7 | oh, wait | 22:21 |
| sjmc7 | that’s a static parameter | 22:21 |
| sjmc7 | you can’t set it on a running index | 22:21 |
| sjmc7 | but there’s no reason we couldn’ t do it at creation time. -1 the patch and i’ll modify it | 22:21 |
| sjmc7 | that’ll allow loads more options, should’ve thought of that before | 22:22 |
| TravT | okay | 22:29 |
| TravT | FYI, i'm putting up a patch on horizon to register the OS::Neutron::FloatingIP labels. | 22:30 |
| *** sjmc7 has quit IRC | 23:09 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!