15:00:55 <frankbrockners> #startmeeting OPNFV BGS weekly team meeting
15:00:55 <collabot> Meeting started Mon Jul 20 15:00:55 2015 UTC.  The chair is frankbrockners. Information about MeetBot at http://wiki.debian.org/MeetBot.
15:00:55 <collabot> Useful Commands: #action #agreed #help #info #idea #link #topic.
15:00:55 <collabot> The meeting name has been set to 'opnfv_bgs_weekly_team_meeting'
15:01:15 <pbandzi> #info Peter Bandzi
15:01:18 <frankbrockners> #topic roll call
15:01:55 <frankbrockners> #info Frank Brockners
15:02:02 <lmcdasm> #info Daniel Smith
15:02:19 <trozet> #info Tim Rozet
15:06:02 <frankbrockners> looks like we're a small group again today
15:06:12 <lmcdasm> aye
15:06:48 <frankbrockners> #topic agenda bashing
15:06:53 <lmcdasm> shall we wait a bit more (say to 10 after the hour) and see if others pop in?
15:07:14 <frankbrockners> #info draft agenda: https://wiki.opnfv.org/meetings/bgs#july202015
15:07:39 <frankbrockners> IMHO we can get started with a topics we can already handle
15:07:42 <lmcdasm> go fot it
15:07:44 <lmcdasm> for*
15:08:02 <frankbrockners> we just have a quick status check in on the agenda
15:08:09 <frankbrockners> should we start with the LF lab reconfig?
15:08:14 <radez> #info Dan Radez
15:08:32 <frankbrockners> #topic LF lab reconfig status (to allow for sequential deployment with different deployment tools)
15:08:48 <frankbrockners> pbandzi - could you give us a quick update?
15:08:59 <pbandzi> yes
15:09:59 <pbandzi> #info we have ucs script which reconfigure ucs network config for fuel or for foreman on one pod. All nodes (including jumphost) are restarted after the change
15:10:52 <pbandzi> #link here it is: https://github.com/pbandzi/lab-reconfiguration
15:11:04 <pbandzi> i put it to github for now
15:11:36 <frankbrockners> #info pbandzi will check the script into OPNFV git
15:11:58 <pbandzi> #info basiclay it changes vnic template assocaition
15:12:02 <pbandzi> end
15:12:02 <trozet> pbandzi: so when do we start using this script and convert to 1 pod?
15:12:05 <frankbrockners> pbandzi - did you try out the script on the LF lab?
15:12:25 <pbandzi> #info script was tried also in LF lab on Friday
15:12:32 <pbandzi> #info seems it is working
15:13:48 <pbandzi> trozet: we can start to use it as soon as possible.
15:14:08 <trozet> pbandzi: so then the JJB jobs will need to be changed to point to use a single jumphost
15:14:32 <trozet> pbanzi: and you mentioned the mac addresses may change ( as a one time thing) so the inventory files will need to be modified
15:14:46 <trozet> does that capture all the changes needed?
15:15:22 <pbandzi> trozet: yes. mac address was probably changed when i started to use vnic templates. they should remain now the same.
15:15:25 <lmcdasm> does anyone (besides Fatih), know how to change the jobs scripts in the jenkins boxs
15:15:33 <lmcdasm> and setup the time cycle to switch back and forth , etc?
15:15:33 <frankbrockners> do we need any associated changes of the LF gateway (I don't assume so, but just asking)
15:16:17 <frankbrockners> lmcdasm - i believe trozet could do the JJB changes as well
15:16:18 <trozet> lmcdasm: i know how to change the JJB now
15:16:18 <pbandzi> don't need changes to LF gateway now. both installers will use same LF gateway...
15:16:25 <lmcdasm> great.
15:16:33 <trozet> its pretty easy actually, all just yaml
15:16:40 <frankbrockners> so let's info the requirements
15:16:46 <trozet> im sure any of us could do it
15:17:06 <lmcdasm> cool.. i was just wondering about the inclusion of the autodeployer as well.. no sweat
15:17:18 <frankbrockners> #info JJB jobs need to be changed - and yaml config files need to reflect new mac addresses
15:17:56 <frankbrockners> can we schedule a day/time to perform the changes and give things a try?
15:18:58 <pbandzi> for me it can be done tomorrow...
15:19:13 <pbandzi> and time - what is suitable for others
15:20:18 <trozet> pbandzi: I'm flexible
15:20:46 <pbandzi> trozet so let's say tomorrow when you arrive to the office
15:20:53 <trozet> pbandzi: ok sounds good
15:20:58 <trozet> only one thing
15:21:10 <frankbrockners> we need to announce this to others as well...
15:21:10 <trozet> we will need a committer on octopus to merge in JJB changes
15:21:30 <frankbrockners> so let's agree on a time
15:21:50 <pbandzi> Tim what time you are in office?
15:22:16 <frankbrockners> dradez is a committer on octopus. dradez could you help?
15:22:45 * trozet didnt know radez was a committer there :)
15:23:00 <trozet> pbandzi: I will be in around 9am EDT
15:23:03 <frankbrockners> ups - got nick wrong
15:23:07 <radez> sure I can help merge
15:23:48 <frankbrockners> should we do 10am EDT? that is the time of the TSC :-)
15:24:28 <pbandzi> ok 10 EDT = 4pm CET
15:24:34 <trozet> ok
15:24:39 <trozet> so which pod will be used?
15:24:45 <pbandzi> pod2
15:25:13 <pbandzi> but in fact it doesn't matter,
15:25:52 <frankbrockners> #info change to sequential deploy is planned for 10am EDT = 4pm CET on Tuesday, Jul/21
15:26:15 <trozet> so is the plan here we switch the installers to POD2
15:26:15 <frankbrockners> #info POD2 will be used for deploy/test from thereon
15:26:24 <trozet> then we dismantle pod1 and break it up into several virtual pods
15:26:57 <frankbrockners> yup - but we should give it a day or two before we do this...
15:27:08 <frankbrockners> let's have at least 2 successful cycles
15:27:44 <frankbrockners> any volunteer who'd announce the planned work to opnfv-tech-discuss?
15:28:05 <trozet> ok so at that point then, can the same script be used to create a pod with 1 server and 1 nic?
15:28:16 <trozet> from the servers left in pod1
15:28:45 <pbandzi> trozet: it can be changed manualy. but if we need a script for it i can make it. or make current more robust
15:29:15 <trozet> pbandzi: well im just thinking if we made pod2 interchangeable to installers, shouldnt we do the same for the new pods created from pod1?
15:29:49 * frankbrockners will write the email - telling folks about the upcoming change
15:30:25 <frankbrockners> trozet: IMHO we'd need a script which performs the network config corresponding to the needs of a specific install.
15:30:52 <frankbrockners> that might mean that the network conf script takes the "installer-config.yml" as input, or?
15:31:00 <trozet> frankbrockners: dont we already have that with the pbandzi script that reprovisions now?
15:32:01 <pbandzi> trozet: sort of. currently script changes network for 6 nodes
15:32:40 <trozet> pbandzi: so your current script must change the network config, from tagged to untagged right, depending on installer?
15:32:41 <frankbrockners> how about we start with the "fixed 6 node pod" script for now
15:32:42 <pbandzi> but if there are reuirements to make it more universal, it might be changed
15:32:55 <frankbrockners> we can make things flexbile in another iteration
15:33:17 <pbandzi> trozet: exactly
15:33:17 <trozet> ok so how about after the pod 2 changes are stable, we create a static 1 pod virtual deployment
15:33:30 <trozet> then pbandzi or somebody can update the script to work on that setup
15:34:08 * frankbrockners likes this plan
15:35:00 <pbandzi> trozet: do you mean script somethink like "give me one node - if available for virtual deployment"
15:35:38 <frankbrockners> #info once setup on POD2 is stable - some servers on POD1 will be repurposed.
15:35:45 <trozet> pbandzi: yeah if virtual, make one node pod with 1 NIC configured
15:35:54 <pbandzi> trozet: ok
15:36:13 <frankbrockners> #info This will include a single server setup deployment for installers (will require further extensions to the current network re-config script by pbandzi)
15:36:15 <trozet> at least thats the case for Foreman, im sure Fuel will be 1 server for virtual, but NIC setup may slightly differ
15:38:48 <frankbrockners> looks like we're done with this topic - and have next steps defined
15:39:00 <frankbrockners> let's move on
15:39:03 <frankbrockners> #Brief status (progress towards Arno SR1)
15:39:04 <lmcdasm> yes tim, thats corrent
15:39:12 <lmcdasm> correct*
15:39:12 <frankbrockners> #topic Brief status (progress towards Arno SR1)
15:39:49 <frankbrockners> Any updates worth mentioning on Fuel or Foreman?
15:40:08 <trozet> #info CI is passing again on Foreman :)
15:40:24 <frankbrockners> You've seen on the alias that folks start to talk about SR1 - so we're likely going to be asked for a date soon
15:40:35 <trozet> #info some major changes were done to deploy and the repos are migrated to genesis now, so we no longer maintain the github bgs_vagrant
15:40:36 <lmcdasm> ive been asked for content lists by a couple people
15:40:45 <lmcdasm> not sure where to point them to at this point since people are gone
15:40:47 <lmcdasm> for summer
15:42:35 <frankbrockners> trozet: do we have everything in opnfv git by now?
15:43:15 <trozet> frankbrockners: the puppet modules are still being used from github, need to migrate those
15:43:31 <trozet> #info will raise JIRA issue to migrate puppet modules (last migration to genesis repo)
15:43:36 <frankbrockners> ok - thanks trozet
15:43:51 <trozet> maybe that should have been #action
15:46:39 <frankbrockners> as long as we have the things captured, we're good :-)
15:47:05 <frankbrockners> do we still see things on track for having SR1 in September?
15:47:14 <trozet> yeah
15:47:19 <frankbrockners> cool
15:47:53 <frankbrockners> lmcdasm - any thoughts?
15:48:16 <lmcdasm> im sorry, but i dont know jonas plan for SR1 content, so i cant really say much :)
15:48:28 <frankbrockners> ok
15:48:51 <frankbrockners> so seems like we're done. anything else?
15:49:27 <frankbrockners> if that's not the case, we can close the meeting ....
15:50:00 <trozet> i think we are done
15:50:13 <frankbrockners> Thanks everyone!
15:50:18 <frankbrockners> #endmeeting