18:59:55 #startmeeting Weekly Poppy Meeting 18:59:56 Meeting started Thu Oct 30 18:59:55 2014 UTC and is due to finish in 60 minutes. The chair is amitgandhinz. Information about MeetBot at http://wiki.debian.org/MeetBot. 18:59:57 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 19:00:00 The meeting name has been set to 'weekly_poppy_meeting' 19:00:06 #topic RollCall 19:00:16 \o 19:00:20 o/ 19:00:31 o/ 19:00:40 What does this sign o/ mean ? 19:00:55 its a face and one hand up waving 19:00:56 its a little guy waving his hand 19:00:57 o/ 19:01:04 o/ 19:01:12 tonytan4ever: must be left handed =P 19:01:15 OK so I wasn't wrong, I am just waing my left hand 19:01:19 \o/ 19:01:22 o/ 19:02:00 ok looks like most of us are here 19:02:01 malini is super happy 19:02:20 I was just letting folks know that I have two hands 19:02:23 #topic last week tonight with amitgandhinz 19:02:41 #link http://eavesdrop.openstack.org/meetings/poppy_weekly_meeting/2014/poppy_weekly_meeting.2014-10-23-19.00.html 19:02:47 no action items 19:02:48 tht toic sounds very scary 19:02:50 that was easy 19:03:29 malini: its a play on the hbo show last week tonight with john oliver 19:03:42 #topic blueprints 19:03:51 #link https://blueprints.launchpad.net/poppy 19:03:57 obulpathi: dns driver 19:04:21 i started it 19:04:30 have basic code 19:04:33 slow or good progress? 19:04:42 good progress :) 19:04:54 but it is coming slowly 19:04:54 cool 19:05:03 thats a contradiction lol 19:05:11 slow and steady 19:05:12 obulpathi: caching-rules 19:05:19 have not started it yet 19:05:42 once I get the patch merged, I will pick it up 19:05:47 ok 19:05:56 migui does not seem to be around today 19:06:15 tonytan4ever: purge-content 19:06:22 It's under review 19:06:29 More reviews are welcome 19:06:30 are ppl reviewing it? 19:06:51 hmm it looks like its ready to merge 19:07:00 oh wait wrong patch ehhe 19:07:31 https://review.openstack.org/#/c/129674/ 19:07:39 you have a -1 from malini to work on 19:07:48 got one comment to address 19:07:55 Should be able to address quickly. 19:07:57 tht was moreof q uestion for everybody 19:08:13 But we can dicsuss it in #openstack-poppy 19:08:18 ok 19:08:26 or in open discussion 19:08:30 amitgandhinz: queue-driver 19:08:42 we need to ban tonytan4ever's space bar 19:08:55 I banned it. 19:08:58 queue-driver is being dev'd by the team this hackday hehe 19:08:58 hhaha 19:09:01 thtwillbebad 19:09:22 obulpathi: patch-service 19:09:37 there are few final details to be ironed out 19:09:48 so its still in the same state 19:09:58 ok 19:10:21 tonytan4ever: set-restrictions 19:10:31 that one is under review as well. 19:10:43 Could use more reviews on that one. 19:11:04 yup. i had some minor nits, but please others review it also 19:11:17 * amitgandhinz the patch, not my nits =P 19:11:18 ok 19:11:39 ok on to bugs 19:11:42 #topic bugs 19:11:53 #link https://bugs.launchpad.net/poppy 19:12:23 i see 2 new ones 19:12:25 one sec 19:12:39 ? 19:12:51 amitgandhinz: I need you to assign the akamai-driver bp to me before we jumping into bugs. 19:12:58 I updated apiary for this one https://bugs.launchpad.net/poppy/+bug/1386315 19:12:59 Launchpad bug 1386315 in poppy "Undocumented "name" property in POST /services error response JSON" [Undecided,New] 19:13:27 done 19:13:32 kk 19:13:38 ok bugs... 19:14:10 can that be closed now malini? 19:14:22 yes 19:14:41 done 19:14:42 https://bugs.launchpad.net/poppy/+bug/1386141 19:14:45 Launchpad bug 1386141 in poppy "Create Flavor - limit value not stored in Cassandra" [Undecided,New] 19:14:57 this looks like a duplicate of https://bugs.launchpad.net/poppy/+bug/1381114 19:14:59 Launchpad bug 1381114 in poppy "Create Flavor - Flavor Limits not validated or saved" [Medium,Confirmed] 19:15:18 no.. 19:15:23 phew..tht was close 19:15:50 well it is :/ 19:15:57 :'( 19:16:08 I lost a new bug.. 19:16:10 * amitgandhinz phew! 19:16:13 lets close the new one 19:16:16 which one can we close 19:16:17 ok 19:16:53 next time I will get a really NEW one 19:17:19 not on my watch ! 19:17:26 =P 19:17:41 any thing else on bugs? 19:17:44 tht is why I test at night BWAHAAHAAA 19:18:09 nope 19:18:21 #topic Paris 19:18:31 one more bug: https://bugs.launchpad.net/poppy/+bug/1384271 19:18:32 Launchpad bug 1384271 in poppy "List Services - NON ASCII limit causes poppy to return 500" [Medium,Confirmed] 19:18:34 this one is fixed 19:18:40 #link https://etherpad.openstack.org/p/poppy-design-session-paris 19:19:04 obulpathi: is that merged? 19:19:22 yes sir 19:19:23 also in the patch commit message you can put Closes Bug: 12345 and it updates the bug 19:19:42 ok, will do 19:19:54 updated. 19:19:57 ok paris... 19:20:11 so i think we finalized the agenda for the design session 19:20:33 wooot! 19:20:36 basically do a 10m intro on poppy (basically the slides i will do in the brown bag) 19:21:10 so that will be about 15 min all up from intro to slides 19:21:27 then i want to spend the majority of the time talking about how poppy can integrate with swift 19:21:32 learn about the challenges 19:21:45 what features we need to support etc 19:21:59 and hopefully come up with a list of things for the poppy team to do and maybe the swift team to do 19:22:38 any comments on this? 19:23:28 i think between all the logging work, getting production ready, and swift integration, that will take us to May for the L summit 19:23:29 looks like a good plan 19:23:55 megan_w_: comments? 19:24:01 I am fine with this plan too. 19:24:16 i thikn this looks good 19:24:29 i think people will be ready for custom ssl, though 19:24:35 probably 19:24:39 we're already hearing it from our sales reps 19:24:45 many people consider CDN useless without it 19:24:50 but i think thats a discussion to be had directly with cdn providers 19:25:00 rather than other openstack teams 19:25:05 agreed 19:25:19 just made the comment since you said those items would take us through may 19:25:32 we may need to find room for ssl in the roadmap before then 19:25:35 im sure there will be more but im just looking at priority list 19:25:44 ya 19:25:45 unless you think ssl is more important than swift 19:25:56 i don't 19:26:00 ok 19:26:41 alright.....anything else to discuss about paris? 19:26:50 cheese? 19:27:06 im going to try to stand at the expo booth on monday, and then design and brownbag on tuesday 19:27:11 and then cheese and wine on wednesday 19:27:18 sounds like a good week 19:27:23 represent us well 19:27:30 and take business cards! 19:27:45 ah yes, must use them up 19:27:59 * amitgandhinz takes them out of drawer and dusts of the webs 19:28:11 yeah, i figured that might be the case :) 19:28:21 he literally did it 19:28:25 :D 19:28:26 lol 19:28:28 amitgandhinz, I mean 19:28:59 ok, moving on 19:29:03 #topic open discussion 19:29:12 31 min remaining to chat about whatever 19:29:26 or we could get back to working =P 19:29:30 so talking about ssl 19:29:34 tonytan4ever had something with spaces 19:29:39 are there are plans to look into keyless ssl 19:29:48 ? 19:29:49 thats the SNI stuff? 19:29:50 Let's chat about one comment to my patchset 19:29:59 https://review.openstack.org/#/c/129674/7/poppy/manager/default/service_async_workers/purge_service_worker.py 19:29:59 ok SSL first 19:29:59 SNI? 19:30:28 http://en.wikipedia.org/wiki/Server_Name_Indication 19:31:00 yes, SNI is becoming a standard poor-mans ssl 19:31:05 and it's plenty for a lot of cusotmres 19:31:09 http://aws.amazon.com/cloudfront/custom-ssl-domains/ 19:31:16 i'd like for us to investigate it 19:31:20 not sure if the SNI and key less ssl means the same 19:31:21 https://www.cloudflare.com/keyless-ssl 19:31:34 if we can get SNI done quickly, it buys us more time for custom SSL certificates 19:32:37 with keyless SSL it sounds like something the CDN providers have to support 19:32:49 since the traffic is NOT going through the operators 19:32:55 amitgandhinz: is that the same as SNI? 19:33:13 it looks similar 19:33:14 maybe someone should spike this 19:33:23 i think SNI is slightly different 19:33:42 keyless is where cloudfront encrypts, decrypts, and reenvrypts the traffic 19:34:04 yes, keyless required CDN providers to support it 19:34:08 SNI is a shared IP with a cert on it 19:34:25 SNI is basically a shared cert, right? 19:34:31 ya 19:34:36 as you just said.. 19:34:48 its more of a shared ip 19:34:52 ..sounds like both require provider support 19:34:52 with multiple certs on it 19:34:56 yup 19:35:16 at the end of the day, poppy should be just enabling exisitng features 19:35:25 and making that provisioning easier 19:35:54 so poppy could be responsible for sharing the cert keys etc (via barbican?) and then the rest is on the providers 19:36:07 i am unclear about what "the rest" is 19:36:08 i think this requrires more spiking though 19:36:12 agreed 19:36:18 agreed 19:36:25 i can start conversations with some of the providers if you'd like 19:36:30 you can give me the action 19:36:31 "the rest" is using the cert, encrypting the traffic, etc 19:36:57 #action megan_w_ to start investigation SNI and Keyless SSL 19:37:17 ok, lets move on to tony's patch 19:37:44 This is the link:https://review.openstack.org/#/c/129674/7/poppy/manager/default/service_async_workers/purge_service_worker.py 19:37:58 see malini's comment inbetween 19:38:06 i think i agree with malini 19:38:13 purge all should be intentional 19:38:47 what if we say the purge url has to be provided and purge all happens if they pass in '/' 19:38:50 make it ?url=all ?? 19:39:02 nah 19:39:10 what if they have a folder called all lol 19:39:14 :D 19:39:19 url=/ ? 19:39:25 thats what im thinking 19:39:28 * ? 19:39:29 I didnt like the url part either 19:40:04 what if they leave the url param out and provide something else instead 19:40:29 ?url=.... will purge a url. ?all=true will purge all 19:40:51 good idea 19:40:54 I like this one. 19:40:55 since anything can be in url 19:40:58 all=True 19:40:59 me too 19:41:35 seems like the right format 19:41:48 what is the full url? 19:41:58 DELETE /services/{service_name}/assets{?url} 19:42:20 ok 19:42:26 so we all agreed? 19:42:33 +1 19:42:43 I'm witing this down! :) 19:42:50 I agree. 19:42:56 we agree too easily :d 19:43:20 #agreed purge all requires a ?all=True parameter instead of ?url=blank 19:43:28 now is it true or True 19:43:34 yeah 19:43:42 tht is test case for me :D 19:43:54 i like lowercase 19:44:10 can it be either? 19:44:13 lowercase if good in urls 19:44:23 why cant it be either? 19:44:28 lowarecase is good in urls 19:44:31 it can, but lets document it in lowercase 19:45:10 cool 19:45:11 +1 19:45:17 any other topics to discuss/ 19:45:18 got it! 19:46:09 none from me.. 19:46:15 do well in Paris amitgandhinz! 19:46:42 thanks! 19:46:47 just don't hold the mic all ghetto like kurt did 19:46:53 hahahaha 19:46:55 you'll do great 19:47:05 wear your chapeau! 19:47:09 lol 19:47:11 i feel i need to buy some beats by dre now 19:47:21 aaaaand we're off 19:47:28 #endmeeting