Hackability Code Yellow Meeting Notes - October 1

[Karen Grunberg]

Action Items:
AI: chase, Adam, Julie will follow up on long term plans
AI: Ojan documentation on sheriff-o-matic and trooper-o-matic
AI: Jam some documentation on what we did, post-mortem
AI: Karen to link docs to wiki page
AI: Folllow up with Armand on whether he can work on
AI: Julie to let us know what Emil’s findings are and long term plans in this area


Notes:
* Karen’s been emailing managers and here are some issues that have been raised:
   - no long term owner
    - need more people reviewing CLs
    - lack of documentation

* We talked to Annie - lack of communication, no visibility (will be sending more notes, like this one, yet!!)

* We have things that have no long term owner - like Alan’s patch from last night that showed up several issues immediately , we need owners for monitoring the monitors

Pillar Updates:
CQ Matches Waterfall - Linux ASAN trybot made progress, sorted out swarming, we only have one config, updated to match waterfall, increased experiment to 50%. Talked to CQ Flakiness this week so we know where to focus on this week. Sergey will work with Sheng on CQ Flakiness and Pawel - Sergiy will work on CQ Matches waterfall.

CQ Flakiness - 7.0% this week, highest flaky bot is 4.8% (went down from 10). How much should the bot contribute to this (should be ~1.2% and not 7, so Sergey is going to look at this discrepancy

Cycle Times: CQ Cycle time 57mins, 90% is 2.2 hours, Working on main tree to get it to use swarming so bots can be faster. We need the push speed faster in Android bots (Armand is working on this.)

Tree Open: Fixed two really bad correctness issues on sheriff-o-matic. Sheriffs using it and filing bugs, yey! 

Prevent Relapse - Mostly discussions with people trying to recruit folks