13:01:22 <jonasbjurel> #startmeeting Fuel weekly sync
13:01:22 <collabot> Meeting started Thu Feb  4 13:01:22 2016 UTC.  The chair is jonasbjurel. Information about MeetBot at http://wiki.debian.org/MeetBot.
13:01:22 <collabot> Useful Commands: #action #agreed #help #info #idea #link #topic.
13:01:22 <collabot> The meeting name has been set to 'fuel_weekly_sync'
13:01:40 <jonasbjurel> #topic Roll-call
13:01:42 <mskalski> #info Michal Skalski
13:01:55 <jonasbjurel> #info Jonas Bjurel
13:01:57 <ethfci> #info Ferenc cserepkei
13:02:04 <s_berg> #info Stefan K. Berg
13:02:11 <jonasbjurel> Hi everyone
13:02:19 <s_berg> Hi!
13:02:24 <ethfci> Hi
13:02:43 <mskalski> Hi, gtm today or irc only ?
13:03:05 <ethfci> gtm not started....
13:03:14 <jonasbjurel> mskalski: irc only, they took all our furnituers in our meeting rooms :-(
13:03:42 <jonasbjurel> We're moving to another building soon
13:03:46 <mskalski> ok :)
13:04:08 <s_berg> We still have our desks though. At least for now.
13:04:16 <ethfci> Jonas how many time we have left until final Brahma?
13:04:38 <jonasbjurel> ethfci: ~20 days
13:04:53 <jonasbjurel> #topic Brahma status
13:05:04 <z0d> #info Peter Barabas
13:05:11 <z0d> can't make it to GTM today, sorry.
13:05:16 <ethfci> i waana die ... its mission impossible :D
13:05:18 <z0d> oh, ok
13:05:31 <z0d> I was reading it backwards
13:05:43 <ethfci> ;D
13:05:43 <jonasbjurel> #info: Looks like things are converging
13:05:54 <jonasbjurel> What issues do we have left?
13:05:59 <s_berg> ethfci: Please stay with us. Banging the head in the wall is usually enough. :)
13:06:08 <ethfci> develop a fuel plugin
13:06:36 <ruijing> Question, what's target release day?
13:06:52 <jonasbjurel> zod update us on the latest deploy work
13:07:19 <z0d> jonasbjurel: I'm running the deployment as we speak
13:07:38 <z0d> if it succeeds, we can merge my changes
13:07:42 <s_berg> z0d: So what was this Keystone issue? Intermittent?
13:07:57 <z0d> s_berg: probably temporary, I'll get back to it
13:08:08 <jonasbjurel> #info Peter has rebased the deployer according to the latest Fuel 8 changes
13:08:22 <jonasbjurel> #info Testing is ongoing.
13:08:33 <z0d> also did some improvements on the deployer with regards to kills and interrupts
13:08:47 <jonasbjurel> zod: info that
13:09:22 <s_berg> I'd suggest we merge that one as well - it will reduce the amount of grief when pressing ^C in the wrong place.
13:09:56 <ethfci> Peter how big is the change?
13:10:09 <z0d> #info did some improvements on interrupt handling (e.g. Ctrl-c). autodeploy should properly clean up now (that is, unmounting the iso)
13:10:19 <jonasbjurel> s_berg: Alread merged to master, waiting merge to stable
13:10:19 <z0d> ethfci: probably 10 lines
13:10:36 <s_berg> https://gerrit.opnfv.org/gerrit/#/c/9137/
13:10:47 <mskalski> Guys, is it still possible to make some minor changes in plugin? ODL Be RC1 is available now, and some things can be also adjusted in manifest to reflect changes in fuel-library from this week
13:11:01 <ethfci> i mean i have a fuel8 capable deploy infra. Shall i validate it once again?
13:11:08 <mskalski> I'm talking about something like this https://review.openstack.org/#/c/276204/1
13:11:24 <mskalski> but I know we are now in code-freeze phase
13:11:35 <jonasbjurel> mskalski: If we test it carefully we might be able to merge it next week.
13:11:40 <z0d> I just run out of free space on Fuel, /var
13:11:59 <z0d> during building the boostrap image
13:12:10 <s_berg> z0d: You're using a 100G for Fuel in the DHA?
13:12:43 <z0d> s_berg: nope, but I've updateing it, thanks
13:12:48 <mskalski> jonasbjurel: ok understand, I currently testing this in my lab
13:12:53 <ethfci> mskalski: wanna have a fuel plugin adds extra features to horizon
13:13:14 <ethfci> is it hard to do in Fuel8
13:13:16 <ethfci> ?
13:13:25 <z0d> do you want me to make fuel 100G in the autodeployer as well?
13:13:56 <z0d> in reap.py specifically
13:14:03 <jonasbjurel> zod I guess so, but that is defined in DHA right?
13:14:11 <z0d> jonasbjurel: yes
13:14:13 <ethfci> 1 am having 64G fuel, 128G compute, controller atm
13:14:21 <mskalski> ethfci: its depend what you want to achieve, but I would not say that is hard, especially when your changes can be applied in post deployment phase
13:14:22 <jonasbjurel> zod Ahh, for reap - yes
13:14:26 <z0d> jonasbjurel: DHA
13:14:37 <z0d> nevermind my previous line
13:15:37 <mskalski> z0d: how big is a /var partition in your case? like 2 day ago there was commit which change size of this partition from fix 10GB to percentage size of hard disk
13:15:49 <jonasbjurel> mskalski: Where do we stand on off-line deployment and local mirrors, do you know?
13:15:53 <z0d> mskalski: it's 7.1G, but fuel only had 50G
13:15:59 <s_berg> z0d: +1 for reap!
13:16:22 <z0d> ok, will do it
13:17:53 <mskalski> jonasbjurel: I would say we are in good shape here, mirrors are included on ISO, and bootstrap image command use it build images, I also do deployments with those mirrors and it worked
13:18:13 <mskalski> there where initial problems with lack of some packages
13:18:17 <jonasbjurel> mskalski: I understood there was an issue with Fuel1 and local mirrors that had been fixed in master but not in stable/fuel8
13:18:30 <ethfci> jonas according my tests, offline deploy is stable and possible
13:18:30 <mskalski> but I fix this in upstream fuel-mirror repository
13:19:09 <s_berg> Then we should cutover to using deploy with the mirror now I think.
13:19:22 <jonasbjurel> Ok, so what do we need to do to enable off-line deployment?
13:19:53 <s_berg> Just a few lines in dea.yaml for the repos. I have a working dea.yaml for that, we can cut'n'paste.
13:19:53 <ethfci> s_berg - SFC deploys run from the iso only ATM
13:20:40 <mskalski> jonasbjurel: configuration of repositories for env in settings.yaml should point to local mirrors
13:21:20 <jonasbjurel> s_berg: It should go in to dea-base under deploy/config
13:22:03 <s_berg> jonasbjurel: OK, will fix and submit a change.
13:22:51 <jonasbjurel> Next, we discussed to freeze upstream to a commit. I think thats a good idea.
13:23:10 <jonasbjurel> mskalski: Do you think now is the time to freeze?
13:23:27 <z0d> (reap fix in https://gerrit.opnfv.org/gerrit/9267 )
13:23:30 <mskalski> in case of bootstrap image I used local address (127.0.0.1) and this configuration is already applied on our ISO, but for environment configuration we need to configure  addresses to have proper fuel master address
13:23:41 <jonasbjurel> zod: Thx
13:23:59 <s_berg> mskalski, jonasbjurel: We should be safe assuming 10.20.0.2 there?
13:24:15 <jonasbjurel> s_berg: +1
13:25:05 <jonasbjurel> My proposal is to freeze tomorrow, any one against?
13:25:18 <ethfci> stefan emailed to you a dea i am using/developed
13:25:39 <z0d> jonasbjurel: +1 for freeze
13:25:39 <s_berg> jonasbjurel: +1, let's then freeze on the commit ids for a verified deploy.
13:25:47 <mskalski> jonasbjurel: probably tomorrow will be hard code freeze for fuel, after that only critical patch will be merged
13:26:01 <s_berg> Great timing then!
13:26:07 <jonasbjurel> mskalski: Then the timing looks right
13:26:24 <mskalski> I will check if this fix https://bugs.launchpad.net/fuel/+bug/1540690 is included in stable/80
13:27:07 <mskalski> sorry not this one, this is already there
13:27:14 <mskalski> this one: https://bugs.launchpad.net/fuel/+bug/1540890
13:27:27 <jonasbjurel> mskalski: I dont think we use RADOS GW in our deployments.
13:27:42 <ethfci> mikhal, so ceph will work again?
13:28:20 <mskalski> ceph worked all the time I think, the problem was connected to RadosGW
13:28:37 <jonasbjurel> mskalski: Then no big deal.
13:29:01 <mskalski> how about Ceph RBD for volumes ?
13:29:05 <DanSmithEricsson> mornin all
13:29:19 <DanSmithEricsson> #info DanielSmith
13:29:26 <jonasbjurel> morning Daniel, did you have cofee yet
13:29:38 <mskalski> ah sorry I mean do we use ceph for store images ?
13:29:43 <DanSmithEricsson> its brewing .. why.. should i look to up the dosage
13:29:44 <DanSmithEricsson> ?
13:29:56 <DanSmithEricsson> (havdnt looked at mails yet either - so be gentle).
13:30:01 <jonasbjurel> :-)
13:30:20 <jonasbjurel> mskalski: RBD for volume is used
13:30:38 <jonasbjurel> Ok lets move forward
13:30:58 <jonasbjurel> Im working on docs, should be ready in a day or two.
13:31:19 <DanSmithEricsson> you gonna redo the II as well with the steps for 8
13:31:20 <mskalski> jonasbjurel: then I would wait until this bug https://bugs.launchpad.net/fuel/+bug/1540890 will be closed, I think it will be today
13:31:26 <DanSmithEricsson> or should i add the new stuff?
13:31:54 <jonasbjurel> Can everyone look at READMEs for ther respective directory.
13:33:04 <jonasbjurel> DanielSmithEricsson: Im redoing it, this time with pictures.
13:33:10 <s_berg> Pushed https://gerrit.opnfv.org/gerrit/#/c/9269/ for turning on repo mirror.
13:33:35 <s_berg> ethfci: Thanks, our changes were identical! Following the same excellent info from mskalski I presume.
13:34:07 <jonasbjurel> Can I ask everyone to create Jira tickets/bugs for all issues we find/know. I will need that for the release notes
13:36:33 <mskalski> jonasbjurel: if odl-l3 only work with vxlan segmentation is this a issue? In plugin I not allow to enable odl-l3 when vlan segmentation is used
13:38:55 <jonasbjurel> mskalski: No problem, we dont use vlan with ODL
13:39:37 <mskalski> ok, it is only about l3, odl-l2+vlan is possible
13:41:12 <jonasbjurel> mskalski: That is OK
13:41:49 <jonasbjurel> Anything more
13:41:50 <mskalski> also I want to help with creating scenarios for deploying ODL, is there place where I can see examples how I can run deploy.py with those new scenarios mechanism?
13:42:25 <jonasbjurel> mskalski: you run it from ci/deploy.sh
13:43:17 <s_berg> There's a README in the ci directory.
13:43:40 <jonasbjurel> ./deploy -h should guide you.
13:43:52 <mskalski> ok will check this, thanks!
13:44:25 <ethfci> mskalski i assume deploy.sh is run by a jenkins slave
13:46:04 <mskalski> ethfci: yeap but for now I only used my local dea and dha files, and I want to test how this jobs in jenkis deploy env with scenarios
13:47:07 <jonasbjurel> Anything more we should discuss now?
13:48:17 <jonasbjurel> I want to thank you all for your excellent work. It looks we will make it again!
13:48:55 <jonasbjurel> Talk to you next week!
13:48:59 <ethfci> Thanks Jonas and have a great weekend
13:49:14 <z0d> jonasbjurel: have a nice weekend
13:49:16 <jonasbjurel> ethfci: Same to you
13:49:23 <ruijing> nice weekend
13:49:36 <ruijing> I will not attend the meeting for holiday:)
13:49:38 <jonasbjurel> zod and all: Have a great weekend
13:49:43 <jonasbjurel> #endmeeting