Wednesday, 2016-04-20

« Tuesday, 2016-04-19 Index Thursday, 2016-04-21 »
*** shu-mutou has joined #openstack-searchlight00:05
*** yingjun has joined #openstack-searchlight00:31
openstackgerritOpenStack Proposal Bot proposed openstack/searchlight: Updated from global requirements  https://review.openstack.org/30777800:42
*** bpokorny has quit IRC01:07
*** yingjun has quit IRC01:13
*** yingjun has joined #openstack-searchlight01:54
*** lakshmiS has quit IRC02:19
*** bpokorny has joined #openstack-searchlight03:40
*** ekarlso has quit IRC04:11
*** bpokorny has quit IRC04:16
*** ekarlso has joined #openstack-searchlight04:25
*** bpokorny has joined #openstack-searchlight05:25
*** bpokorny has quit IRC05:56
openstackgerritHiroyuki Eguchi proposed openstack/searchlight: Document that initial sync is mandatory  https://review.openstack.org/30815206:48
*** yingjun has quit IRC07:08
*** yingjun has joined #openstack-searchlight07:12
*** david-lyle has quit IRC08:59
*** david-lyle has joined #openstack-searchlight09:00
*** yingjun has quit IRC09:46
*** yingjun has joined #openstack-searchlight12:49
*** sigmavirus24_awa is now known as sigmavirus2414:15
openstackgerritTravis Tripp proposed openstack/searchlight: Document that initial sync is mandatory  https://review.openstack.org/30815214:30
*** yingjun has quit IRC14:38
*** yingjun has joined #openstack-searchlight14:38
openstackgerritOpenStack Proposal Bot proposed openstack/searchlight: Updated from global requirements  https://review.openstack.org/30777814:42
*** yingjun has quit IRC14:43
openstackgerritMatt Borland proposed openstack/searchlight-ui: Move search bar out of table  https://review.openstack.org/30654015:00
*** bpokorny has joined #openstack-searchlight15:00
*** yingjun has joined #openstack-searchlight15:23
*** nikhil_k has joined #openstack-searchlight15:38
*** nikhil has quit IRC15:40
*** yingjun has quit IRC15:40
*** yingjun has joined #openstack-searchlight15:41
*** yingjun has quit IRC15:46
*** lakshmiS has joined #openstack-searchlight15:49
*** lakshmiS has quit IRC16:15
*** lakshmiS has joined #openstack-searchlight16:15
*** bpokorny_ has joined #openstack-searchlight16:18
*** bpokorny has quit IRC16:22
openstackgerritSteve McLellan proposed openstack/searchlight: Make func test indexing more consistent  https://review.openstack.org/30598516:40
*** bpokorny_ has quit IRC17:06
*** bpokorny has joined #openstack-searchlight17:07
lakshmiShi TravT18:15
TravThey lakshmiS18:16
lakshmiSfor resource type info in reference id fields in facets;18:16
lakshmiSare you expecting the resource name field for every id fields. lets say for nova servers, on image id field should it be like this18:17
lakshmiS"OS::Nova::Server": [18:17
lakshmiS        {18:17
lakshmiS            "type": "string",18:17
lakshmiS            "name": "image.id",18:17
lakshmiS            "options": [],18:17
lakshmiS            “resource_type” : “OS::Glance::Image”18:17
lakshmiS        }18:17
lakshmiS}18:17
* TravT thinking.18:19
TravTif we do it in facets, then that would work well I think.18:20
lakshmiSso in the UI, you will lookup image details based in the id and resource type if user needs to expand the id field?18:21
TravTyeah, something like that... but in the UI this is also useful from a static context.18:22
TravTlet me explain a bit18:22
TravTif I just do a freeform search across all resources18:22
TravTI may get a dozen results from different resource types18:22
TravTthe data may include various IDs18:23
TravThowever, i don't request the facet list until we want to limit search to a specific type18:23
TravTthis goes along the lines of a conversation sjmc7 and i had the other day about returning the total18:23
* sjmc7 hides18:24
TravTif we only give the resource_type info when request the facet list for a type and the fields.18:24
TravTwe would have to request facets for all resource types to get this static info.18:24
TravTbut this does seem pretty natural place to put the info.18:25
lakshmiSand for parent info, the same would be like18:27
lakshmiS "OS::Swift::Container": [18:27
lakshmiS              {18:27
lakshmiS            "type": "string",18:27
lakshmiS            "name": "id”,18:28
lakshmiS            “_parent”: “OS::Swift::Account”18:28
lakshmiS        }18:28
lakshmiS]18:28
sjmc7just a heads up i’m also modifying the facet code so one of us will have a time rebasing, but what i’m doing is probably less urgent18:28
TravTlakshmiS: FYI: these logs: http://eavesdrop.openstack.org/irclogs/%23openstack-searchlight/%23openstack-searchlight.2016-04-19.log.html18:29
TravTi will be right back. need to step away for about 2 mins18:29
TravTlakshmi _parent: actually doesn't work right?18:35
TravTit makes an assumption of _parent18:35
TravThowever, that is dynamic in plugin18:36
TravTbase.py get_parent_id_field18:36
lakshmiSi was exploring elasticsearch to see if we can add metadata to mapping definitions18:37
TravTand then just expose the mapping?18:37
TravTi wondered about that as well18:37
lakshmiShmmm...18:37
sjmc7i don’t think so18:38
lakshmiSyou will not get options18:38
lakshmiSor anything you need to refine18:38
lakshmiSi was more interested in asissting the facet generation logic to get info from mapping18:39
TravTahh18:39
TravTmaybe putting this in facets is okay as a starting point....18:40
lakshmiSwhats the alternative? plugin_info?18:41
TravTyes, because its static.18:41
sjmc7it was maybe a mistake not to separate the static and dynamic stuff to begin with18:41
sjmc7facets is a mishmash of both18:41
*** sigmavirus24 is now known as sigmavirus24_awa18:45
TravTlakshmiS:18:50
TravTfor parent18:50
TravTmaybe in the parent field, you'd just have18:50
TravT"parent": "OS::Swift::Account"18:51
TravTor actually parent: true18:51
lakshmiSwithout the type info?18:51
lakshmiSoh18:51
TravTbecause if you had “resource_type” : “OS::Swift::Account”18:51
TravTand "parent": true18:51
TravTyou'd know that fields type and know that it is a parent to this resource type18:52
lakshmiSright18:52
lakshmiSso if we add parent field to facets, what happens to plugin_info result.18:54
TravTjust a sec, i'm playing with some queries...18:55
lakshmiSi mean is that useful anywhere else?18:55
*** sigmavirus24_awa is now known as sigmavirus2418:56
lakshmiSFYI: I will be adding meta info on mappings using _meta feature in ES : https://www.elastic.co/guide/en/elasticsearch/reference/1.4/mapping-meta.html18:57
TravToh, nice18:58
TravTfor simple parent child things, if i know that a type has children (even though I don't know what they are), i could issue a has_parent query and then just render all results blindly.19:04
TravTthat won't be as performant...19:04
TravTif I know the children of a type, then i can issue a query for each type and just match on the parent field.19:05
TravTlet me go try a few things in the UI.19:08
lakshmiSi will quickly grab lunch..19:09
lakshmiSback19:37
openstackgerritZhiQiang Fan proposed openstack/searchlight: use thread safe fnmatch  https://review.openstack.org/30855720:06
TravTlakshmiS: i'm still working in the UI on this.20:14
lakshmiSTravT: if you think it will be easy to do hangout to go check this out, let me know(i may understand it with a better context)20:33
openstackgerritSteve McLellan proposed openstack/searchlight: Add totals to facets  https://review.openstack.org/30858220:42
openstackgerritMerged openstack/searchlight: Document that initial sync is mandatory  https://review.openstack.org/30815220:59
*** bpokorny_ has joined #openstack-searchlight21:02
*** bpokorny has quit IRC21:07
lakshmiSFYI: code coverage for searchlight is at 43%21:33
openstackgerritMerged openstack/searchlight: use thread safe fnmatch  https://review.openstack.org/30855721:38
openstackgerritMerged openstack/searchlight: Making searchlight-manage python-3 compatible  https://review.openstack.org/30091121:45
*** lakshmiS has quit IRC21:53
*** nikhil_k is now known as nikhil21:58
*** sigmavirus24 is now known as sigmavirus24_awa22:15
*** krotscheck is now known as krotscheck_dcm22:57
openstackgerritOpenStack Proposal Bot proposed openstack/searchlight: Updated from global requirements  https://review.openstack.org/30777823:18
*** bpokorny_ has quit IRC23:25
*** bpokorny has joined #openstack-searchlight23:26
*** lakshmiS has joined #openstack-searchlight23:41
openstackgerritTravis Tripp proposed openstack/searchlight-ui: Two Column Default for Search result Drawer  https://review.openstack.org/30869623:43
« Tuesday, 2016-04-19 Index Thursday, 2016-04-21 »

Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!