07:01:49 #startmeeting Second Arno adjusted target release plan planning meeting :D 07:01:49 Meeting started Thu Apr 23 07:01:49 2015 UTC. The chair is ChrisPriceAB. Information about MeetBot at http://wiki.debian.org/MeetBot. 07:01:49 Useful Commands: #action #agreed #help #info #idea #link #topic. 07:01:49 The meeting name has been set to 'second_arno_adjusted_target_release_plan_planning_meeting__d' 07:01:55 #topic roll call 07:02:03 #info Chris Price 07:02:06 #info Uli 07:04:07 #topic Release readiness update 07:04:25 #link https://wiki.opnfv.org/releases/arno/releasereadiness release readiness task list 07:04:49 #info we discussed last night beginning to put dependancies in this task list, I also had an action to update the opnfvdocs pieces. 07:05:03 #info these did not start last night after the meeting, we can do so today. 07:05:36 #info we agreed to use "themes" as a way or organising activities into dependancies, rather than trying ot use the numbered list. 07:05:54 #info Frank Brockners 07:06:12 Can you explain "themes" 07:06:14 #info we also agreed to provide some project planning on the wiki and develop that during today. 07:06:29 #info Themes = high-level categories/groups 07:06:39 #info have a look at the wiki 07:06:49 OK 07:07:03 #info The themes we to be idetifyable activities that might contain sub-tasks that are easily refferable as depenadancies on the other items (like a story) 07:07:08 #info first column is the category - things that naturally belong together 07:07:35 yup - epic -> userstory is a good analogy 07:07:40 #info a theme is like SYS.READY which requires all system readiness activities to be complete prior to the theme itself being ready 07:08:21 ok. Just feared there is some software called "themes" ..... :D 07:08:27 hehe 07:08:48 #info we additional articulated the task dependancy that forms the process required to achieve Arno completion 07:08:51 #info Time to release readiness: max{TEST.READY, SYS.READY} + max{SYS.AUTOMATED,TEST.EXEC} + SYS.RELEASE + DOCS.READY 07:09:26 #info this provides us with a way of combining tasks and pending time into an equation that gives an indicative release date 07:09:32 + 1-2 days risk buffer 07:09:36 * ChrisPriceAB very engineering, thanks frank 07:10:10 well the buffer is in the final, or is it in the people estimating time for a task buffer+buffet and an added buffer = a long time in my experience 07:10:42 #info we still have not begun to work with functest, morgan is back in the office today. 07:11:13 * ulik Poor guy 07:11:33 #info and there are themes/categories which have an "unknown" time for now - true for SYS.RELEASE, SYS.READY and TEST.READY 07:11:38 #info in parallel jose_lausuch will work on Tempest automation, pbandzi will work on robot automation, the app deploy teams are working in parallel. 07:11:50 #info no timelines have been developed for these activities yet. 07:12:45 #action all project leads to ensure we have time estimates for remaining tasks and that the dependancies are defined on other tasks 07:13:11 #info this will provide a method of estimating parallelism and the critical path for the remaining project activities 07:13:52 questions? 07:14:28 #info we need to work in parallel on the WiKi during the day and will agree between the project leads on the final pre-TSC meeting today on the proposed plan 07:14:40 Can we give each task a number (or a 3-5 char name) just to use it in the dependencies 07:14:58 we should use the "theme" names. numbers move around too much 07:15:09 the list will probably shuffle during the say 07:15:11 day 07:15:25 I meant fix numbers..... 07:15:41 We might need dependencies inside a theme 07:16:01 Or is that too fine granular? 07:16:09 on the TSC call later today: do we feel like we're able to present a rough estimate for the additional time required? 07:16:18 Yeah, for theme internal dependencies I'm ok for that to handled in such a way. I hope there are not internal theme dependancies that we need to expose. 07:16:24 We have to give a date. 07:16:32 'But tbh, I expect the project leads have that under control. 07:16:46 ulik - IMHO we don't need those dependencies within a team - it is clear for those working on the tasks and for others it does not matter 07:17:04 frankckbrockners, that is what we have to work toward today. 07:17:32 ulik: BGS can't give a date until we have technical certainty on SYS.RELEASE and we know what TEST.READY is 07:17:38 I doubt we will be scientifically accurate. But we need to be confident enough that we can tackle thing in a given timeframe. 07:17:41 i.e. we don't even know what will be tested 07:18:16 yes we have cyclic dependancies, test depends on bgs depends on test which both depend on specific ci activities... 07:18:32 we need to manage those and focus on task completion tracking in the project plan. 07:19:17 hopefully Jose get's things on track today for functest. We gave him a set up lab and said get test running :D 07:19:35 this should begin to feed-back probable times to competion. 07:19:46 hopefully in time for the meeting today :D 07:20:06 Ok... 07:20:25 So you want every PL to fill column 7 in each line with his project in col.5 ? 07:20:31 #topic activities to be completed prior to Final pre-TSC Arno planning meeting. 07:20:53 And just put there something like SYS:DOCS as dependency. 07:21:05 #info fill in column 7 (dependancies) where needed to be drawn out 07:21:27 #info fill in column 6 (time to completion) with estimates on task completion timelines 07:21:50 #info only add dependancies were there cannot be parallelism (that includes resources) 07:22:39 #info for tasks "almost complete" where we will be troubleshooting when a dependant task tries to use it, it may be better to dfactor that cost in to the dependant task. opinions? 07:23:28 #action functest team to complete the project with best known estimates 07:23:45 #action opnfvdocs team to compete actions outlined in the previous meeting 07:24:39 #action PL's to work on the project plan text on the wiki. We can add a description of our tasks in simple language there which we can edit after lunch today into a singular Arno text 07:24:58 the plan shuld be described in text under franks equation. 07:25:06 we all good? anything else? 07:25:27 I'll follow up with trevor once he is up and morgan once I can get hold of him 07:26:26 ulik, frankckbrockners: shall we close? 07:26:36 ok 07:26:40 ChrisPriceAB: we can close... but 07:26:50 thanks guys, onward and upward 07:26:52 #endmeeting