19:02:00 <catherineD> #startmeeting refstack 19:02:01 <openstack> Meeting started Mon Jun 29 19:02:00 2015 UTC and is due to finish in 60 minutes. The chair is catherineD. Information about MeetBot at http://wiki.debian.org/MeetBot. 19:02:02 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 19:02:04 <openstack> The meeting name has been set to 'refstack' 19:02:04 <sslypushenko__> o/ 19:02:11 <Rockyg> o/ 19:02:20 <catherineD> Roll call 19:02:35 <davidlenwell> o/ 19:03:06 <catherineD> hello every one .. Chris won't be able to make today's meeting .. 19:03:14 <pvaneck> o/ 19:03:24 <catherineD> please see agenda #link https://etherpad.openstack.org/p/refstack-meeting-15-06-29 19:03:47 <catherineD> #topic Relocate RefStack Project 19:04:27 <sslypushenko__> With Rob's vote, I think Infra wins) 19:04:44 <catherineD> We were tied during last week's vote ... but I think zehecle: broke the tie 19:04:50 <catherineD> sslypushenko__: yes 19:05:32 <davidlenwell> did rob vote? 19:05:35 <catherineD> davidlenwell: do you know what is the next step 19:05:57 <sslypushenko__> Actually, I think there is no big real difference. So let it be "Infra" 19:06:02 <catherineD> not officially by he sent me a note and I forward to everyine ... 19:06:36 <catherineD> davidlenwell: I sent to your gmail ... could you check? 19:06:37 <davidlenwell> making the move is easy .. 19:06:57 <zehicle> o/ 19:07:02 <davidlenwell> I have no email from you catherineD 19:07:16 <catherineD> sorry my typo is really bad today .. 19:07:24 <davidlenwell> The issue is that I have a full plate until tuesday when I go on vacation until monday 19:07:35 * zehicle confirms +1 on Infra 19:07:39 <davidlenwell> So I can take care of getting us moved into infra.. but It will have to be next week 19:08:07 <catherineD> zehicle: thx for your vote ... 19:08:31 <Rockyg> That should be fine. I might be able to start the ball rolling. I'm doing a couple of other repos right now. 19:08:51 <catherineD> #agreed Moving Refstack under Infra's namespace 19:09:20 <krotscheck> o/ 19:09:28 <catherineD> #action davidlenwell: to take care of the next steps to move Refstack into Infra 19:10:05 <davidlenwell> cool .. like I said .. when I get back next week I'll handle it 19:10:15 <catherineD> davidlenwell: thx 19:10:30 <catherineD> #topic Infra Hosting 19:10:43 <davidlenwell> our spec was merged.. https://review.openstack.org/#/c/188207/ 19:11:27 <davidlenwell> from what I understand we are waiting on a review from jeblair on this .. https://review.openstack.org/#/c/182437/ 19:11:50 <catherineD> davidlenwell: hogepodge: Good job! Thank you! 19:12:27 <catherineD> #link https://review.openstack.org/#/c/188207/ merged !!! 19:12:46 <catherineD> Wating for jeblair's +2 on #link https://review.openstack.org/#/c/188207/ 19:13:23 <catherineD> anything else to discuss on this topic? 19:13:31 <Rockyg> catherineD: got the last review number wrong... 19:13:48 <davidlenwell> yeah .. those last two links are the same 19:13:50 <catherineD> Rockyg: thx 19:14:42 <catherineD> Wating for jeblair's +2 on #link https://review.openstack.org/#/c/182437/ 19:15:20 <catherineD> #topic Set up eavesdrop for refstack channel 19:15:48 <catherineD> This is Chris' action item ... will check with him next week ... 19:16:43 <catherineD> #action Check with hogepodge: on status next week 19:17:17 <catherineD> #topic OpenStackID integration 19:18:01 <catherineD> sslypushenko__: I guess https://review.openstack.org/#/c/176335/ need bebase 19:18:32 <sslypushenko__> catherineD I got email from you and I tried to reproduce that error but failed ( 19:19:08 <catherineD> sslypushenko__: thx ... we can discuss more on #refstack .. 19:19:10 <sslypushenko__> I will send you email with details 19:19:21 <sslypushenko__> agreed 19:20:02 <sslypushenko__> Patch rebased 19:21:06 <catherineD> We just need to resolve some of the issues ... but I have no doubt that we will merge it .... let's disucss some of the topics related to it ... 19:21:16 <sslypushenko__> pvaneck Did you try this patch? 19:21:38 <davidlenwell> its just a matter of gettings the silly tests to pass .. 19:22:01 <pvaneck> i tried it with the run-in-docker script, but had to modify the osid_endpoint in the refstack.conf.tmpl 19:22:19 <pvaneck> it worked okay with my local openstackid 19:22:33 <sslypushenko__> Good news) 19:23:10 <sslypushenko__> I will think how to provide osid_endpoint little bit accurate 19:23:11 <catherineD> good news indeed .... let me compare my env with pvaneck:'s 19:24:12 <sslypushenko__> oh! This might help 19:24:12 <catherineD> Currently, we identify a cloud by using Keystone UUID as its cpid (cloud provider id) ... 19:25:31 <catherineD> sslypushenko__: pvaneck: let me know if you are ready to discuss next topic cpid ... 19:25:48 <sslypushenko__> I'm ready) 19:25:56 <catherineD> thx 19:26:20 <catherineD> we need to have admin credential to fecth keystone UUID ... 19:26:44 <sslypushenko__> But don't have solution) 19:27:05 <sslypushenko__> Without admin creds keystone api is really small 19:27:45 <catherineD> zehicle: davidlenwell: Rockyg: that means that Refstack will always require admin credential .. not for test ... but for fetching cloud id 19:28:16 <davidlenwell> to clarify refstack-client will need admin creds 19:28:21 <davidlenwell> which we decided early on was fine 19:28:36 <zehicle> yy, that's a reality we have to accept 19:28:47 <davidlenwell> the refstack api / database does not store any creds to clouds 19:28:55 <zehicle> would be nice for it to be addressed over time so that end users could submit 19:29:10 <sslypushenko__> It is fine for now. But in future it will be not fine, I think 19:29:17 <davidlenwell> we have a lot of work to do with q/a before that can be a reality 19:29:38 <catherineD> This means that Refstack will always need to be tested on-premise .. we will not be able to graduate to have a centralized test 19:29:39 <davidlenwell> I'm not saying we shouldn't do it .. im just saying that this is not new information 19:29:40 <Rockyg> +1 19:30:28 <davidlenwell> centralized testing would be nice .. but I don't think its realistic 19:30:55 <davidlenwell> how can you test admin api's from outside the admin network.. you just can't .. there are more problems here than just creds 19:31:17 <davidlenwell> logistically the test needs to be ran inside an enviornment 19:31:22 <Rockyg> amen to that 19:31:34 <davidlenwell> we can revist tcup as a solution 19:31:40 <catherineD> So I take that we all agree that we will consider this topic but will defer it to later 19:31:45 <davidlenwell> but I don't think trust is really a problem here 19:31:48 <sslypushenko__> I don't think that getting cpid with refstack-client is really necessary 19:32:07 <catherineD> sslypushenko__: how would we identify a cloud? 19:32:09 <davidlenwell> with signed results it isn't as nescesary depending on what you plan on doing with the data 19:32:22 <sslypushenko__> User can specify it 19:32:27 <davidlenwell> yeah 19:32:29 <davidlenwell> that works too 19:32:35 <davidlenwell> good thinking sslypushenko__ 19:32:43 <catherineD> so user will pass in keystone uuid? 19:33:03 <davidlenwell> use "can" 19:33:07 <sslypushenko__> Some thing like name, will be good enought 19:33:20 <davidlenwell> but the refstack client needs admin creds anyways .. so I don't understand the point of this discussion 19:33:33 <sslypushenko__> or something more unique 19:33:57 <davidlenwell> okay.. I have to run now.. sorry gang.. hit me up in #refstack if anyone wants to chat about this further 19:34:09 <sslypushenko__> davidlenwell q/a are moving fast) 19:35:06 <catherineD> davidlenwell: with the latest change using yaml file ... I believe a lot of the test results that Chris received were collected without admin credential .. we can confirm with Chris on that ... 19:36:32 <catherineD> so do we agree to use cpid for now or we decide that all data upload will be signed and user needs to pass in cloud ID (name or UUID) 19:37:02 <sslypushenko__> cpid for now is ok 19:37:40 <catherineD> sslypushenko__: +1 19:37:46 <sslypushenko__> if something will change we will adjust to changes easly 19:38:03 <pvaneck> yea, that's fine for now 19:38:50 <catherineD> Rockyg: zehicle: your thoughts? 19:39:29 <Rockyg> agreed. We can "enhance" later 19:39:48 <zehicle> sorry - have to catch up 19:41:02 <zehicle> my understanding is catherineD is right - if you know the CPID then you have rights for that CPID 19:41:26 <zehicle> at some point, we'll have to be able to cluster them into user accounts 19:41:49 <catherineD> #agreed Using Keystone UUID as cpid for now. Thus refstack-client still require admin credential. 19:42:30 <catherineD> #agreed Revisit the CPID tonpic in the next cycle 19:42:59 <zehicle> is there a refstack mid-cycle planned? 19:43:43 <catherineD> not yet .. but that is one of today's topic ... 19:44:41 <catherineD> let's continue on the user authentication topic ... 19:44:53 <zehicle> kk, thanks 19:45:07 <catherineD> Currently, on refstack.net, the results are displayed anonymously .. once user can log in with their own ID .. we need to add a page that displayed their own data ... 19:45:53 <sslypushenko__> sure, I'm working on that 19:46:11 <catherineD> great ... thx sslypushenko__: 19:47:04 <sslypushenko__> I'm working on endpoint for updating user profile. It will allow to upload user pubkey 19:47:38 <sslypushenko__> So after that we can add filter to UI for user's own tests 19:49:24 <catherineD> #action sslypushenko__: working on user;s own Refstack test result pages 19:50:12 <pvaneck> so to associate data with accounts, all youd have to do is use the --sign arg in refstack-client? 19:50:34 <sslypushenko__> yeap 19:50:42 <catherineD> we have 10 mins ... let's just discuss the 2 items in the Open Discussion ... that means we skip topic 5, 6, 7 for now? 19:51:24 <catherineD> #topic Open discussion 19:51:34 <catherineD> Do we need a Refstack mid-cycle f2f? 19:52:02 <sslypushenko__> I think, uit would be helpful 19:52:06 <pvaneck> weren't you at least planning on meeting with Rocky and David in the bay? 19:53:10 <Rockyg> yep. but I can't remember for what 19:54:27 <catherineD> Rockyg: I think we were to meet for deciding where to relocate Refstack ... but we have made our decision today .. 19:54:57 <catherineD> I am thinking of a f2f to discuss action items #5 in https://etherpad.openstack.org/p/refstack_f2f_may_2015 19:54:57 <Rockyg> Hmm. I thought it was to write something??? 19:55:25 <pvaneck> lol would have to go back and check logs 19:56:04 <catherineD> I will check the log but I think it is for relocation of Refstack ... 19:56:33 <pvaneck> it's okay. We can still schedule another f2f if need be for other items 19:56:56 <catherineD> let discuss f2f next week ... 19:57:20 <catherineD> #action Discuss midcycle f2f meeting next week 19:57:49 <catherineD> 3 mins for next dicussion : Do we want to submit a speaker session for Tokyo summit? 19:58:42 <pvaneck> do you have any topics youd wish to talk about? 19:59:34 <pvaneck> deadline i think is the 15th of July 19:59:47 <catherineD> I am thinking of we should have a session to discuss the new features added to refstack and refstack-client since Kilo cycle 20:00:11 <Rockyg> ++ 20:00:11 <catherineD> we need to end meeting 20:00:25 <catherineD> #endmeeting