15:01:35 #startmeeting OPNFV Pharos 15:01:35 Meeting started Wed Sep 23 15:01:35 2015 UTC. The chair is trevor_intel. Information about MeetBot at http://wiki.debian.org/MeetBot. 15:01:35 Useful Commands: #action #agreed #help #info #idea #link #topic. 15:01:35 The meeting name has been set to 'opnfv_pharos' 15:01:50 #info Trevor Cooper 15:01:57 #info Michael Wynne 15:01:58 #info Chris Price 15:02:10 #info Bin Hu 15:02:17 #info Agenda proposal ... 15:02:44 #info 1. Status of current activities 15:02:55 #info a) MAAS PoC 15:03:04 #info b) CI resources 15:03:17 #info c) Creating template examples 15:03:30 #info 2. Milestones 15:03:48 #info 3. Rls C plans 15:03:57 #info a) Project dependencies 15:04:06 quick question: is there an audio track for this meeting or is it all in IRC? thanks. 15:04:21 #info b) Readiness of first activities/sprints 15:04:35 #info c) Issues and risks 15:04:53 JoeKidder: not really 15:05:05 #info Comments / questions on agenda please 15:05:26 we can set one if for those that can’t IRC and we can help them bridge questions into IRC 15:05:32 but it’s some small effort 15:05:57 we can't use teh GTM since other meetings are happening now 15:06:05 nope - IRC is fine - I just wasn't sure if I was missing a great soundtrack - I'll just watch;) 15:06:25 yeah! MaaS POC is first on the agenda - !!! i’m happy about that - 15:06:39 agenda looks good to me - no comments 15:06:56 #topic MAAS PoC 15:07:10 Who can give a readout ... Iben? 15:07:18 Nrinder? 15:07:23 is narindergupta here? 15:07:25 Narinder (sorry) 15:07:32 we have some progress 15:07:45 the rc and cc machines are setup 15:07:51 iben: yes i am here 15:08:16 now waiting on some “pod” machines to be built out - expect eta this week or early next - this is issue on the spirent side 15:08:21 right now waiting for jump host so we can register 15:08:52 Is this in Spirent or Intel lab (or both)? 15:08:53 also need to get the pod 3 setup in intel lab - narindergupta - any news on that? 15:08:58 its both 15:09:05 spirnet lab from my update 15:09:26 for intel lab i got the new pod access pod6 yesterday for MAAS POC 15:09:27 Question ... what is the differnce beteen PoC in Spirent and Intel lab or is it a duplicate activity? 15:09:32 we had preliminary meeting with dell lab and they are hoping to get in on the poc too but have some requirements 15:09:37 For my part I owe Narinder a lab spec for POD6. 15:09:44 our goal for poc is to show multisite capability for maas 15:10:02 so we need to have more than one and less than five labs 15:10:08 centally controlled 15:10:13 correct 15:10:21 yes but haev to start with one and then built it out? 15:10:30 that plus the ability to recycle a pod from one config to next in less than an hour 15:10:52 we already have this runnign in the intel pod 5 - but not centrally to other remote labs 15:11:03 the issue with intel lab is the lack of public ip addresses 15:11:20 So focus should be central (LF) part .... yes? 15:11:26 iMSWynne: so we will use intel pod 6 for maas poc - not intel pod 3? 15:11:38 iben: correct 15:11:45 INtel POD 3 is VSPERF 15:11:55 trevor_intel: focus is any central server - can start in spirent or somewhere “easy” then move to other location when ready 15:12:24 it’s not easy for us to get servers and network changes in the LF lab - plus - that is for prodution purpose - not POC e xperiemnt - right? 15:12:32 Narinder: did the firewall rules work, or do we still have an issue? 15:12:34 sorry for typos... 15:12:40 iben: so which lab is the central one for first phase of teh PoC? 15:12:45 spirent 15:13:08 Intel is the slave and Spirent is central? 15:13:19 regional control is on spirent 15:13:46 and mutiple cluster controller spready across spirent and intel will have cluster controllers 15:13:47 Ok and tehn you plan to migrate to LF lab? 15:14:05 correct then regional controller will get migrated to linux foundation 15:14:09 slaves in both intel and spirent - not slave but we call them cluster controllers - a cc can control multiple pods 15:14:11 goti it 15:14:21 so many cc in many labs 15:14:34 governed by one master rc or regional controller 15:14:53 This is our first milestone: MAAS PoC  complete with recommendations re. infrastructure management system 15:14:58 once we get the rc working will with one or more cc we will move the rc to lf after a demo review by the community 15:15:08 Comments on first milestone? 15:15:23 another aspect of the poc is authentication - we need to have some sort of integration evaluated 15:15:35 we are doing research on ldap type auth now 15:16:01 soon we will need to start thinking abotu federated identity with all these labs and people sharing resources 15:16:05 When will we reach first milestone? 15:17:11 narindergupta: Is first milestone stated correctly? MAAS PoC  complete with recommendations re. infrastructure management system 15:17:35 narindergupta: What is target for completion? 15:17:44 good question - it depends on the resources and lab setup velocity - i’m shooting for next week - but narindergupta might know more 15:18:04 trevor_intel: we are trying if we have maas cc controller in labs then most likely next week 15:18:25 iben correct 15:18:39 narindergupta: all complete ... including recommendations documented? That seems incredibly agressive 15:19:05 I haven't seen any diagrams explaining scope yet 15:19:22 there is some documentation on the setup - but the recommendations might be simple bullet list of what to do 15:19:53 trevor_intel: we have diagram let mw share with you 15:19:54 "Come up with a one/ two pager for the POC and a drawing of what the end to end setup looks like" 15:20:32 trevor_intel: https://wiki.opnfv.org/get_started/lflab_maaspoc 15:21:07 We need trackign in Jira ... will make it easy to follow progress (I think all permission issues are solved?) 15:21:11 the concept is written already - now we “prove” it - right? POC? 15:21:26 yes - JIRA permissions are working better now - thanks! 15:21:38 yes it is 15:21:54 i’m certain we can reproduce the initial MaaS and JOID setup for this POC - the risk for me is the extra things we want to accomplish 15:22:01 like authentication 15:23:02 like resource reservation 15:23:59 like deploying fuel/foreman through MAAS and fuel/foreman successfully deploy Arno release using the topology created by MAAS. 15:24:27 like defining minimum resources needed in LF for MAAS to "work" across 10 (say) labs ... need to reserve hardware or can we live with a VM (with bridges on existing controller)? 15:24:30 is fuel foreman part of milestone 1? 15:24:43 deploying fuel/foreman through MAAS we can demostrate 15:24:44 Jira says 2 topologies 15:24:58 I think that means 2 deploers? 15:25:02 deployers 15:25:03 yes - today we have 2 - odl controller and contrail controller 15:25:27 these are the 2 topologies we will deliver next week in multi-site setup 15:25:29 What deployers? 15:25:51 er? maas is today running joid 15:26:03 we can do it with joid definetely but for fuel and foreman only Arno images 15:26:04 so for this POC a assume we will continue with that 15:26:22 expanding this to other installers has not been done but can be added as milestone 2 15:26:31 would that be okay? 15:27:14 If you can expand the one/two pager to explain these details it will go a long way IMO 15:27:47 I think that scope is still not clear to those of us who are not doing the work :) 15:27:58 okay - thanks! this is good feedback 15:28:08 Also please comment on all the realted Jira epics 15:28:31 Thanks all MASS contributors! 15:28:35 i’m still learning too - so just passing on MY understanding - the proof is in the pudding - i am expecting we can pull it off - we shall see. 15:28:36 MAAS 15:28:48 lets move on 15:29:07 #topic CI resources 15:29:27 iMSWynne: anything to report? 15:29:47 or anybody else? 15:30:14 Yes, the systems are ready in POD4. Received the connections strings from Aric. 15:30:26 Installing todayy. 15:30:54 iMSWynne: great! 15:31:16 Looks like we can move to next topic :) 15:31:34 #topic Template examples 15:31:46 Is there anybody from orange? 15:31:52 Orange 15:32:33 I still couldn't assign tasks ... Cyril, Olivier, Bertrand have not logged into Jira yet. 15:32:55 I will follow up with them 15:33:01 next topic ... 15:33:23 #topic Milestone planning 15:33:51 #link https://etherpad.opnfv.org/p/pharos_brahmaputra 15:34:15 Look at bottom of Etherpad 15:35:02 I need help/input to get the miletones in shape ... and some dates 15:35:19 How can we do this? 15:35:35 We can go through each one now? 15:35:42 and get comments 15:35:50 ... lets try that ... 15:35:53 ok? 15:36:13 Or other ideas? 15:36:34 ok, I have to drop in a few minutes for another meeting though 15:38:09 #info 1. MAAS PoC  complete with recommendations re.     infrastructure management system 15:38:21 I think we covered that 15:38:36 #info . Templates created 15:38:47 #undo 15:38:47 Removing item from minutes: 15:39:01 #info 2. Templates created 15:39:37 These are Lab specs, complinace checklist for dev and CI labs, lab access and user guides 15:41:10 #info 3. Management system setup and documented 15:42:15 #info 4. Establish community lab model for CI with goal to support CI Pipeline for all Rls B installers and include a testing scheme 15:42:41 #info 5. CI lab model implemented in one community lab with documentation 15:43:19 we have a template for the lab spec, right? Has anything else been done with #2? I'd like to know how much time I have 15:44:50 iMSWynne: Teh spec we have was thron together for Rls A 15:44:53 thrown 15:45:16 We need a "proper" spec with templates for each community lab to fill 15:45:25 ok, good. 15:45:35 Orange has been assigned to create an example template 15:46:03 #info 6. Establish community lab model for Dev with goal to support dev and test projects include a testing scheme 15:46:16 #info 7. Lab spec templates filled by participating community labs 15:46:18 i see some people in the etherpad but they are not “named” 15:46:22 they are annoymous 15:46:31 #info 8. Dev lab model implemented in  one community lab with documentation 15:46:53 Where can I find templates 15:47:16 VikramDham: Our job is to create them 15:47:40 And we need owners for the tasks 15:47:46 I like to know what others have in mind for the timeline on #7 15:47:51 Otherwise it will not happen 15:48:09 ok 15:48:50 iMSWynne: We should ask Orange who are assigned initial task of creating example templates 15:49:38 Names are not assigned in Jira yet because assignees haev not logged into Jira (last time I checked) 15:50:25 oops I have to drop in 10 min 15:50:54 #info Trevor requests input on milestones documented on Ehterpad https://etherpad.opnfv.org/p/pharos_brahmaputra 15:51:59 #action ALL pharos committers and lab owners are assigned task to review milestones and update/edit/comment (ASAP please) 15:52:17 I will enter it as a task in Jira 15:52:37 what is the IRC command for assigning ARs? 15:53:08 #topic Rls C plans 15:53:32 I am not too clear on dependencies 15:53:42 Ci has a dependency on Pharos 15:53:55 What about projects Pharos is dependent on? 15:54:34 Also need niput on risks? 15:54:42 inpput 15:55:03 What are the top issues? 15:55:26 #info quick question: how much will the Pharos B Release be different from Arno release http://artifacts.opnfv.org/arno.2015.1.0/docs/pharos-spec.arno.2015.1.0.pdf? 15:56:46 In another word, if we start to build a lab based on Pharos in Arno, how much new effort shall we anticipate? 15:57:22 bin_: expanded scope ... there will be two models 1) CI lab model 2) Dev lab model 15:57:45 sorry I have a hard stop 15:57:55 I assume Arno Pharos is CI Lab model, right? 15:58:20 No problem, and we can continue to discuss on mailing list 15:58:29 bin_: no it was for dev ... all CI resources were in LF lab 15:59:05 I see. Let me write an email to you to further discuss it. 15:59:15 I will fill out Rls milestone C repoort and assign Jira tasks to review and comment on milestones etc. 15:59:56 bin_: I can respond next week ... I am on crital path of an intenal release and swamped 16:00:01 sorry 16:00:06 #endmeeting