*** david-ly_ has joined #openstack-searchlight | 00:04 | |
*** david-lyle has quit IRC | 00:05 | |
*** david-ly_ has quit IRC | 01:14 | |
*** david-lyle has joined #openstack-searchlight | 01:22 | |
*** david-lyle has quit IRC | 01:29 | |
*** itisha has quit IRC | 02:11 | |
*** david-lyle has joined #openstack-searchlight | 03:47 | |
*** akanksha_ has quit IRC | 06:18 | |
*** GB21 has joined #openstack-searchlight | 06:32 | |
*** GB21_ has joined #openstack-searchlight | 06:32 | |
*** GB21_ has quit IRC | 06:33 | |
*** akanksha_ has joined #openstack-searchlight | 08:13 | |
*** nikhil has quit IRC | 08:26 | |
*** nikhil has joined #openstack-searchlight | 08:27 | |
*** nikhil has quit IRC | 09:02 | |
*** nikhil has joined #openstack-searchlight | 09:03 | |
*** sballe has quit IRC | 09:35 | |
*** sballe has joined #openstack-searchlight | 10:22 | |
*** GB21 has quit IRC | 10:35 | |
*** GB21 has joined #openstack-searchlight | 11:29 | |
*** akanksha_ has quit IRC | 12:48 | |
*** itisha has joined #openstack-searchlight | 13:25 | |
*** lakshmiS has joined #openstack-searchlight | 13:59 | |
*** sigmavirus24_awa is now known as sigmavirus24 | 14:05 | |
*** lakshmiS has quit IRC | 14:12 | |
*** GB21 has quit IRC | 14:39 | |
*** lakshmiS has joined #openstack-searchlight | 14:47 | |
*** sjmc7_ has joined #openstack-searchlight | 14:49 | |
*** GB21 has joined #openstack-searchlight | 15:08 | |
*** itisha has quit IRC | 15:51 | |
*** lakshmiS has quit IRC | 17:11 | |
openstackgerrit | Travis Tripp proposed openstack/searchlight: Improve documentation for Liberty release https://review.openstack.org/228664 | 17:11 |
---|---|---|
*** lakshmiS has joined #openstack-searchlight | 17:13 | |
*** GB21 has quit IRC | 17:18 | |
*** GB21 has joined #openstack-searchlight | 17:24 | |
*** lakshmiS_ has joined #openstack-searchlight | 17:56 | |
*** lakshmiS has quit IRC | 17:58 | |
*** itisha has joined #openstack-searchlight | 18:01 | |
itisha | sjmc7, TravT: hey, just saw yesterday's messages. So, no code changes are required anymore right? just some documentation changes? meaning the patch isn't needed anymore right? | 18:13 |
openstackgerrit | Merged openstack/searchlight: Improve documentation for Liberty release https://review.openstack.org/228664 | 18:13 |
sjmc7_ | itisha - yes, i think so | 18:27 |
itisha | hmm, ok, alright :) | 18:29 |
*** sigmavirus24 is now known as sigmavirus24_awa | 18:34 | |
*** sigmavirus24_awa is now known as sigmavirus24 | 18:35 | |
sjmc7_ | sorry :( | 18:35 |
itisha | oh that's completely alright, no problem at all! :) | 18:38 |
*** GB21 has quit IRC | 18:46 | |
*** lakshmiS_ has left #openstack-searchlight | 19:01 | |
*** lakshmiS_ has joined #openstack-searchlight | 19:07 | |
openstackgerrit | Steve McLellan proposed openstack/searchlight: Don't index admin-only fields https://review.openstack.org/233225 | 19:08 |
sigmavirus24 | sjmc7_: don't have time to review it but I looked at the commit message of ^ | 19:23 |
sigmavirus24 | I like the idea of a separate admin-only index | 19:23 |
sjmc7_ | yeah... we were treating it as a semi-security issue | 19:23 |
sigmavirus24 | sjmc7_: that's cool, no worries | 19:24 |
sigmavirus24 | my elasticsearch-fu is super weak though, would admin-only be joinable with the regular index? | 19:24 |
sjmc7_ | unfortunately we discovered it a bit late to implement that | 19:24 |
sjmc7_ | we'd index stuff twice | 19:24 |
sigmavirus24 | ew | 19:24 |
sigmavirus24 | I was worried about that | 19:24 |
sjmc7_ | it's actually the preferred approach | 19:24 |
sigmavirus24 | really? | 19:24 |
sigmavirus24 | interesting | 19:24 |
sigmavirus24 | Like I said, super weak elasticsearch-fu | 19:25 |
sjmc7_ | elasticsearch (+ lucene) is designed in a no-sql kind of way, index it as you want to use it | 19:25 |
sigmavirus24 | I'm more experienced at consuming it than administering it | 19:25 |
sigmavirus24 | Okay | 19:25 |
sjmc7_ | yeah. it's quite common to denormalize | 19:25 |
sigmavirus24 | Would admin-only be a Mitaka thing then? | 19:25 |
sjmc7_ | yeah. i'm too wiped out to reliably get that implemented in the next day or two | 19:25 |
sjmc7_ | but i think probably shortly after the summit | 19:26 |
sigmavirus24 | Yeah I wasn't asking you to do it, sorry if it came off that way | 19:26 |
sjmc7_ | no, it didn't | 19:26 |
sjmc7_ | i wish we could | 19:26 |
* sigmavirus24 is trying to keep abreast of things | 19:26 | |
sigmavirus24 | even if I'm not getting time to work on openstack | 19:26 |
sjmc7_ | yeah. it's a shame we didn't find it last weekl | 19:26 |
TravT | sigmavirus24: the bug report has more on that. | 19:26 |
sigmavirus24 | sjmc7_: nothing's perfect on the first try and it's best that we caught it and can limit the security impact than had we not | 19:27 |
sjmc7_ | yeah | 19:27 |
TravT | we've been off to the side working all night and morning on this one... | 19:27 |
sjmc7_ | it was a foolish mistake in retrospect | 19:27 |
sjmc7_ | but better now than later | 19:27 |
TravT | sjmc7: that isn't how the alias works | 19:27 |
TravT | you index one | 19:27 |
TravT | once | 19:27 |
TravT | then you create separate aliases | 19:28 |
sjmc7_ | oh. ok :) i maybe didn't read what you said properly | 19:28 |
TravT | each alias allow for specifying a filter | 19:28 |
TravT | the filter is matches | 19:28 |
sjmc7_ | ok. i'll need to re-read | 19:28 |
TravT | that affect what is visible in both results and for searching | 19:28 |
TravT | then when the request comes into searchlight | 19:28 |
sjmc7_ | it'll intercept things like querystrings? | 19:28 |
TravT | we evaluate if they are admin or not | 19:28 |
sjmc7_ | if so they've expanded what it can do | 19:29 |
TravT | and we send the request off to the correct alias | 19:29 |
sjmc7_ | since i last used it | 19:29 |
TravT | that's my understanding, anyway | 19:29 |
lakshmiS_ | number of aliases would increase for glance | 19:29 |
sjmc7_ | my only experience with aliasing is to simulate an index-per-user type model | 19:29 |
sjmc7_ | where you apply additional filters | 19:29 |
TravT | yeah, we are all a bit wiped right now | 19:30 |
TravT | but we need to look at the alias for two things | 19:30 |
sjmc7_ | i misread what you said as multiple indices, so i'll reread it :) | 19:30 |
TravT | this and for zero downtime reindexing | 19:30 |
TravT | but let's tackle that another day | 19:30 |
sjmc7_ | a good summit discussion | 19:30 |
TravT | i've almost got the doc update done for the patch above... just a couple more minutes | 19:30 |
openstackgerrit | Merged openstack/searchlight: Don't index admin-only fields https://review.openstack.org/233225 | 19:31 |
lakshmiS_ | we dont have to worry about setting priorities now :) | 19:31 |
TravT | sjmc7: would you be able to now cherry pick the above and propose to stable/liberty? | 19:31 |
TravT | that way we can include in rc2 | 19:32 |
sjmc7_ | yep | 19:32 |
sjmc7_ | did you update .gitreview on stable/liberty? | 19:33 |
sjmc7_ | i'll do that too | 19:33 |
TravT | yeah i did | 19:33 |
sjmc7_ | thanks. that'll save a number of swears in future :) | 19:34 |
sjmc7_ | https://review.openstack.org/233239 TravT | 19:35 |
sjmc7_ | let me know when the doc update's there | 19:35 |
*** lakshmiS_ has quit IRC | 19:35 | |
*** lakshmiS has joined #openstack-searchlight | 19:36 | |
openstackgerrit | Travis Tripp proposed openstack/searchlight: Document unsearchable_fields configuration option https://review.openstack.org/233250 | 19:47 |
TravT | sjmc7_: ^ | 19:47 |
sjmc7_ | looking | 19:47 |
TravT | sjmc7: was https://review.openstack.org/#/c/233239/ cherry-picked? | 19:48 |
sjmc7_ | yeah | 19:48 |
TravT | ok. | 19:48 |
sjmc7_ | why.. does it look weird? | 19:48 |
TravT | for some reason i thought there'd be something in the commit message. | 19:48 |
sjmc7_ | nope | 19:49 |
sjmc7_ | oh.. maybe you're meant to do that. one sec | 19:49 |
sjmc7_ | i'll add a note | 19:49 |
sjmc7_ | i don't think it's automatic | 19:49 |
TravT | (cherry picked from commit fooo) | 19:49 |
TravT | see here: https://review.openstack.org/#/c/229399/ | 19:50 |
sjmc7_ | yeah | 19:50 |
sjmc7_ | well spotted | 19:50 |
TravT | i don't drink energy drinks for nothing | 19:51 |
sjmc7_ | hmm.. it doesn't seem to want to link it | 19:54 |
TravT | change id looks short | 19:55 |
TravT | chould be a full sha | 19:55 |
sjmc7_ | the commit id? | 19:55 |
TravT | like that review above i linked | 19:56 |
sjmc7_ | yeah.. others have the shortened one | 19:56 |
TravT | ... so hungry | 19:56 |
sjmc7_ | i'll give it one more go | 19:56 |
sjmc7_ | ok, that did it | 19:57 |
*** lakshmiS has quit IRC | 20:09 | |
openstackgerrit | Travis Tripp proposed openstack/searchlight: Document unsearchable_fields configuration option https://review.openstack.org/233250 | 20:16 |
TravT | sjmc7^ I just remembered another tidbit on the above. | 20:18 |
TravT | so have another update on the way | 20:18 |
sjmc7_ | my finger was on the +2 button | 20:19 |
openstackgerrit | Travis Tripp proposed openstack/searchlight: Document unsearchable_fields configuration option https://review.openstack.org/233250 | 20:29 |
TravT | sjmc7_: updated... ^ | 20:29 |
sjmc7_ | ah, good addition | 20:30 |
sjmc7_ | will check the results once jenkins finishes, looks good tho | 20:30 |
TravT | david-lyle: sigmavirus24: rosmaita: if you have a chance to look at the above documentation update, it would be appreciated. thanks! | 20:38 |
TravT | i have to run to do some errands, but will address any comments later today. | 20:38 |
*** akanksha_ has joined #openstack-searchlight | 20:42 | |
openstackgerrit | Merged openstack/searchlight: Document unsearchable_fields configuration option https://review.openstack.org/233250 | 20:52 |
*** sjmc7__ has joined #openstack-searchlight | 20:59 | |
*** sjmc7_ has quit IRC | 21:00 | |
sigmavirus24 | ccneill: /goto bops | 21:06 |
sigmavirus24 | ignore that ;) | 21:06 |
*** sjmc7__ has quit IRC | 21:22 | |
*** sjmc7_ has joined #openstack-searchlight | 21:34 | |
*** sjmc7_ has quit IRC | 21:51 | |
*** sjmc7_ has joined #openstack-searchlight | 21:52 | |
*** sigmavirus24 is now known as sigmavirus24_awa | 21:58 |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!