Skip to content
This repository has been archived by the owner on Jul 16, 2020. It is now read-only.

Weekly Meeting 2016 08 25

Jump to bottom
Tim Pepper edited this page Aug 25, 2016 · 6 revisions

Agenda

##Minutes

#ciao-project Meeting

Meeting started by mrkz at 16:04:28 UTC. The full logs are available at ciao-project/2016/ciao-project.2016-08-25-16.04.log.html .

Meeting summary

Meeting ended at 16:59:49 UTC.

Action Items

Action Items, by person

  • UNASSIGNED
    • (none)

People Present (lines said)

  • mrkz (63)
  • markusry (45)
  • leoswaldo (24)
  • kristenc (15)
  • mcastelino (15)
  • albertom (9)
  • obedmr- (8)
  • tcpepper1 (7)
  • sameo (6)
  • ciaomtgbot (3)
  • ciaoslackbridge (1)
  • fuentess (1)
  • _erick0zcr (1)
  • tcpepper (0)

Generated by MeetBot_ 0.1.4

.. _MeetBot: http://wiki.debian.org/MeetBot

###Full IRC Log

16:04:28 <mrkz> #startmeeting
16:04:28 <ciaomtgbot> Meeting started Thu Aug 25 16:04:28 2016 UTC.  The chair is mrkz. Information about MeetBot at http://wiki.debian.org/MeetBot.
16:04:28 <ciaomtgbot> Useful Commands: #action #agreed #help #info #idea #link #topic.
16:04:45 <mcastelino> o/
16:04:48 <mrkz> let's start with our weekly team
16:04:49 <mrkz> o/
16:04:57 <tcpepper1> #rollcall
16:05:01 <tcpepper1> #topic rollcall
16:05:08 <mrkz> #chair tcpepper
16:05:08 <ciaomtgbot> Current chairs: mrkz tcpepper
16:05:09 <tcpepper1> o/
16:05:12 <leoswaldo> o/
16:05:14 <mrkz> #topic rollcall
16:05:38 <markusry> o/
16:05:39 <_erick0zcr> o/
16:05:40 <obedmr-> o/
16:05:54 <mrkz> right, so let's start with our opens please
16:05:57 <mrkz> #topic opens
16:06:01 * tcpepper1 has noe
16:06:03 <tcpepper1> none
16:06:08 <mrkz> any opens you want to bring out the meeting?
16:06:15 <markusry> I have one.
16:06:15 <sameo> o/
16:06:16 <leoswaldo> I have one
16:06:23 <mrkz> markusry: please go ahead
16:06:31 <mrkz> (then leoswaldo's open)
16:06:55 <markusry> It's just about the open PRs.  There are quite a lot.  There's a bit of a backlog due to last week.
16:07:21 <markusry> Do we want to divide up these PRs between the maintainers for reviewing and merging
16:07:22 <markusry> ?
16:08:17 <mrkz> #question There's a lot of Pull Requests open (9), Do we want to divide up these PR between maintainers for review/merge?
16:08:27 <markusry> 463 and 459 seem to be okay and just need to be merged.
16:08:55 <markusry> I'm not sure what the status of 443, 449 and 451 are.
16:09:09 <markusry> I'm already looking at 489 and 488.
16:09:51 <mcastelino> markusry: I need to try #449 on my system to ensure it works on ubuntu... then I can merge it
16:09:53 <mrkz> PR (451) needs rework as some netmasq tests are failing and need to check/fix
16:10:17 <markusry> mrkz: Great.  Could you modify the title temporarily to say don't merge.
16:10:28 <markusry> Make's it easier to filter out.
16:10:43 <mrkz> markusry: just did
16:10:51 <markusry> That just leaves 443
16:11:37 <mrkz> who could review/merge that one?
16:11:40 <tcpepper1> I can review 443
16:11:52 <mrkz> thanks tcpepper1
16:12:14 <mrkz> leoswaldo: is you open related to PR 443 ?
16:12:16 <markusry> I think I can merge 459. kristen's already reviewed it.
16:12:23 <sameo> mrkz: I'll have a look at 488
16:12:33 <leoswaldo> no
16:12:50 <markusry> Could someone merge 463?  It's mine so I don't want to merge myself
16:13:10 <markusry> sameo: Great. I've already made some comments on 488
16:13:24 <sameo> markusry: I just saw that, cool.
16:13:47 <mrkz> so can we say we're dividing those PR reviews/merges? (to answer previous question)
16:13:59 <markusry> Yes.
16:14:11 <mcastelino> sameo: can you try #449 on FC? I wil try it it on Ubuntu
16:14:38 <markusry> Shall I try to make a consolidated list?
16:14:46 <markusry> Of what we just agreed?
16:14:49 <sameo> mcastelino: I'll do so tomorrow.
16:14:59 <mrkz> #info current PR (443, 449, 459, 488) are divided for reviews/merges
16:15:03 <mrkz> thanks folks
16:15:12 <mrkz> let's continue with leoswaldo's open please
16:15:20 <markusry> I still need someone to merge 463
16:15:29 <leoswaldo> ok
16:16:25 <leoswaldo> ciao API docs automation is on its way, right now we are adding a job to the release process, so every time a new release is generated, the docs are updated
16:16:44 <leoswaldo> I've seen kristenc is using her credentials to commit the tags
16:16:50 <leoswaldo> and here my question comes
16:17:12 <leoswaldo> is there a way we can create a general user to commit those changes
16:17:15 <leoswaldo> ?
16:17:33 <kristenc> it will depend on whether intel allows faceless accounts to github.
16:17:49 <kristenc> (to the 01org project)
16:18:31 <markusry> Here's the list
16:18:32 <markusry> 488: sameo, markusry
16:18:32 <markusry> 443: tpepper
16:18:32 <markusry> 449: sameo, mcastelino
16:18:32 <markusry> 459: markusry
16:18:34 <markusry> 489: markusry,tpepper
16:18:36 <markusry> 463: tpepper
16:18:47 <markusry> tpepper: I gave you 463.  I think you looked at it last week
16:19:05 <mcastelino> markusry: I took care of 463
16:19:11 <leoswaldo> why would we neeed a faceless account for that, can't that be on github side ?
16:19:22 <markusry> mcastelino: thanks
16:19:50 <kristenc> pushing the tags requires commit rights to the repo and membership in the org I believe.
16:20:29 <markusry> what exactly do we mean by docs are updated?
16:20:32 <obedmr-> 01org admin should be able to add non-intel (generic github) account
16:20:42 <markusry> Will we be pushing generated html files somewhere?
16:20:51 <mrkz> or just editing the wiki or.. ?
16:20:52 <leoswaldo> markdown  to the wiki
16:20:55 <mrkz> leoswaldo: ^
16:21:07 <leoswaldo> https://github.com/01org/ciao/wiki/Compute-API
16:21:45 <markusry> I see, so each time we change an API the wiki gets automatically updated?
16:21:48 <leoswaldo> that page will be updated every new release if changes
16:21:54 <leoswaldo> yeap markusry
16:22:38 <markusry> I see.
16:22:48 <leoswaldo> question, who is a 01org admin, to expler the non-intel user accout ?
16:22:57 <leoswaldo> explore*
16:23:03 <sameo> leoswaldo: This needs to be handled through a JIRA.
16:24:01 <mrkz> leoswaldo: can we continue with your open after the agenda? (so we can continue with bug scrub as we're running out of time)
16:24:10 <mrkz> looks like it's gonna take a bit more of time
16:24:11 <leoswaldo> ok, iI will file a jira to track it
16:24:18 <mrkz> leoswaldo: thanks
16:24:18 <leoswaldo> sure
16:24:22 <leoswaldo> thanks guys
16:24:34 <mrkz> so if there's no more opens I'd like to move to the bug scrub now
16:25:23 <mrkz> #topic bug scrub
16:25:57 <mrkz> so we have no P1 on the agenda but we have few P2 to check
16:25:58 <mrkz> #link https://github.com/01org/ciao/issues?q=is0X0P+0open+is0X0P+0issue+label0X0P+0bug+label0X0P+0P2
16:26:42 <kristenc> do we not have any p1 bugs? woot.
16:26:59 <mrkz> 1st p2 is #450 "ciao-cli: workload launch cli hangs when CNCI fails to get an IP address" open by mcastelino assigned to albertom
16:26:59 <mrkz> #link https://github.com/01org/ciao/issues/450
16:27:24 <kristenc> p2 bugs are lower priority than p1 features. would there be any updates in reality here to p2 bugs?
16:27:46 <albertom> yes i saw that one and assigned to me
16:27:48 <albertom> will check later
16:27:53 <markusry> My 3 bugs (#343, #17, #16) are all still valid and I haven't started looking at them yet.
16:28:01 <markusry> I think P2 is okay for these.
16:28:55 <mrkz> markusry: so no updates for those?
16:29:07 <markusry> Nope.
16:29:47 <mrkz> right, thanks for the report on those
16:30:01 <mrkz> so we also have a bunch of P3
16:30:01 <mrkz> #link https://github.com/01org/ciao/issues?q=is0X0P+0open+is0X0P+0issue+label0X0P+0bug+label0X0P+0P3
16:31:05 <mrkz> #info no updates on P2 Issues (450, 343, 17, 16)
16:31:42 <markusry> 306, 268, 267, 193, 72 are all ciao-vendor issues.
16:31:47 <markusry> And all still valid.
16:31:50 * tcpepper1 votes for leaving them as-is...effectively backlog
16:32:01 <markusry> I need to have a ciao-vendor day sometime to fix these.
16:32:19 <markusry> Was planning to do it while travelling but got distracted by QMP.
16:32:46 <markusry> They don't affect the running of a ciao cluster
16:32:52 <mrkz> fuentess: any update on https://github.com/01org/ciao/issues/39 ? (e.g. have you tried to reproduce with sameo's comment ? )
16:33:37 <fuentess> mrkz: Haven't tried yet, will try this week and comment there
16:33:50 <mrkz> fuentess: thanks for checking
16:34:54 <mrkz> markusry: any coment also on https://github.com/01org/ciao/issues/11 ? "FULL command should not contain a payload" (low prio, no updates? )
16:35:14 <markusry> No.  It's still valid but benign.
16:35:20 <mrkz> correct
16:35:22 <markusry> Haven't fixed it yet
16:36:05 <mrkz> sameo: I guess no updates also for "ciao-cli: Request to create two instances sometimes results in the creation of a single instance"  ? https://github.com/01org/ciao/issues/61
16:36:15 <sameo> mrkz: No, sorry.
16:36:23 <mrkz> gotcha
16:37:20 <mrkz> so we're just missing 320 from mcastelino
16:37:34 <mrkz> which is unassigned
16:37:42 <mrkz> https://github.com/01org/ciao/issues/320
16:37:56 <mrkz> "networking - firewall unit tests should not panic"
16:38:23 <kristenc> priority seems valid - I think it should be assigned to mcastelino
16:38:41 <mcastelino> mrkz: yes.. and also 487... I need to fix them both
16:39:01 <mrkz> mcastelino: mind you if I assing it to you?
16:39:10 <mrkz> thanks kristenc
16:39:29 <mcastelino> mrkz: go ahead.. will fix them this week...
16:39:36 <mrkz> thanks for the update
16:40:17 <mrkz> so that's all for the bug scrub (unless something is hiding from me)
16:40:48 <mrkz> there's also prior meeting actions listed
16:40:50 <mcastelino> mrkz: when do we discuss the configuration open
16:40:56 <mcastelino> or do it over email?
16:41:59 <mrkz> mcastelino: was moved for next week becasue sameo is @ linuxcon and would like to him to be present for a better discussion (tough, yesterday's talk was very nice)
16:42:27 <mcastelino> makes sense
16:42:33 <mrkz> mcastelino, sameo: https://github.com/01org/ciao/wiki/Weekly-Meeting-2016-09-01
16:42:34 <ciaoslackbridge> <> Pssst! I didn’t unfurl <https://github.com/01org/ciao/wiki/Weekly-Meeting-2016-09-01> because it was already shared in this channel quite recently (within the last hour) and I didn’t want to clutter things up.
16:42:59 <mrkz> right, so anything from this week prior meeting actions left to discuss?
16:43:06 <mrkz> otherwise I'd like to call the meeting
16:44:37 <mrkz> we only have "swagger (5 minutes): status update?  branch merged to master but has CI failures?" and the "easier than easy deploy (10 minutes)"
16:45:26 <mrkz> the 1st one is already being addressed by leoswaldo, and the missing part for 2nd one is being reviewed in https://github.com/01org/ciao/pull/489
16:45:39 <mrkz> so we're covered I'd say, any comments on this?
16:46:01 <obedmr-> leoswaldo: ^^
16:46:43 <leoswaldo> its failing only in 1.7
16:47:04 <markusry> Oh great.  The patch we just merged.
16:47:17 <leoswaldo> I'll take a look at it today, for some reason its converting the  golang.com  to golang_org
16:47:28 <markusry> Do you have a link to the build?
16:47:38 <leoswaldo> I need to figure out why it does that
16:48:07 <mrkz> leoswaldo: ^ ?
16:48:13 <leoswaldo> https://travis-ci.org/01org/ciao/jobs/154570092
16:48:25 <obedmr-> Regarding the keystone and ciao-webui containers they're ready and available at: https://hub.docker.com/r/clearlinux/
16:48:58 <mrkz> #info Kesytone and ciao-webui containers ready and available at https://hub.docker.com/r/clearlinux/
16:49:02 <kristenc> hurray! I was just about to ask about a webui container.
16:49:09 <obedmr-> albertom: could you share the Ansible stuff status?
16:49:10 <markusry> But it's the same problem with that version of swagger that doesn't handle vendoring correctly
16:49:20 <markusry> How does kubernetes do this?
16:49:22 <leoswaldo> and yes, swagger already accepted my PR last week, so we are good to play with swagger master
16:49:28 <kristenc> I had thought that just as ciao could deploy keystone vm it could also deploy the webui once we have public IP merged.
16:49:43 <albertom> well all the progress is in https://github.com/01org/ciao/pull/485
16:50:35 <albertom> as ciao start doing more of the setup by itself, the tasks can be removed from ansible
16:50:43 <albertom> like the cnci image creation for example
16:50:46 <mcastelino> kristenc: markusry: should we deploy the infrastrue VM's /containers directly on the underlay network (like CNCI?)
16:52:19 <kristenc> mcastelino, I think so. I think we should use ansible to "boot strap" ciao, then have ciao deploy itself.
16:52:43 <albertom> kristenc: +1
16:53:25 <mcastelino> kristenc: I already support creation of infrastructure VM's (CNCI being just one of them)... I need to add support for infrastructure containers....
16:53:39 <kristenc> we need to be able to assign public ips to some of these services as well.
16:54:25 <mcastelino> kristenc: if they are on the underlay... then we can do the public IP mappings using the bastion
16:54:27 <kristenc> controller already supports the idea of "internal" workloads (i.e. cnci), I just need to add more of them.
16:54:47 <obedmr-> mcastelino: would launcher be able to assign IPs to containers?
16:55:31 <mcastelino> obedmr-: good point.. it has to be picked from a pool of reserved underlay network IP's
16:55:47 <mcastelino> VMs can DHCP.. containers typically do not
16:56:00 <albertom> kristenc: just be careful to not end in a depdency loop. I saw some folks storing the Active Directory VM in a Storage that needed access to the AD to start serving files :P
16:56:09 <albertom> there was a poeroff and they could start the AD server
16:56:21 <kristenc> albertom, we of course will not design a system like that :).
16:57:13 <albertom> even the greates make mistakes like that one but yeah its a good idea to have ciao deploy keystone itself
16:57:30 <mcastelino> kristenc: let me send out a proposal for "internal" container workloads... we already support "internal" VM workloads...
16:57:50 <kristenc> mcastelino, ok.
16:57:51 <mrkz> folks, we're running out of time, should we add ^ to next week agenda or continue the discussion after the meeting?
16:58:46 <kristenc> we should discuss asap, as we have some urgency on getting our deployment solid i think.
16:58:59 <kristenc> so whenever it makes sense.
16:59:26 <mrkz> right so I'm closing the meeting, but topic is still open, thanks for attending folks!
16:59:28 <obedmr-> for now, ansible is already deploying keystone and ciao-webui
16:59:34 <obedmr-> with docker containers

Weekly Meetings

Development

Architecture

Clone this wiki locally