While bot health is separate from tree health, I'm not sure we really want to tolerate redness because a bot is sick--it'll still keep us from noticing regressions or bustage in a timely fashion. However, a way to visually distinguish between "build failure" and "bot failure" would be nice. "out of disk space" should be easy to detect--is there any reliable way to detect purify running out of RAM, or other bot failure modes we know about?
--Amanda On Fri, Aug 14, 2009 at 3:23 PM, Marc-Antoine Ruel<[email protected]> wrote: > One thing I've been thinking about his adding many more automatic tree > closure on failure. This way the sheriff *have* to react. > > I guess I'll do it sooner than later. > > The downside is when want to tolerate redness; example, the purify bot > currently dies because of lack of memory (yay!) I'll find out a way to > work around that. > > M-A > > On Fri, Aug 14, 2009 at 3:18 PM, Amanda Walker<[email protected]> wrote: >> On Fri, Aug 14, 2009 at 2:28 PM, Peter Kasting<[email protected]> wrote: >>> I'm pretty frustrated with this week's tree. Last Thursday and Friday Tim >>> and I kept *everything* green at almost all times. This week there's been >>> constant redness, huge swaths of orange on the WebKit LayoutTest bots, and >>> the tree's been open much of that time. I have done a few things to help >>> but I feel like I can either let the tree be red or be a full-time sheriff. >>> That's not how it should be. >> >> One thing that my monday/tuesday stint as webkit sheriff brought to >> mind was that perhaps we need to start having weekend sheriffs, or >> somehow reinforce the "even if no sheriff is sitting in IRC, nobody's >> off the hook for redness" message... >> >>> I wrote about what both sheriffs and others need to do at all times a week >>> and a half ago. Maybe nobody read my mail because it was too long, or maybe >>> just nobody cares about the tree being in good shape. It's disappointing. >> >> I don't think "nobody read my mail" or "nobody cares" are accurate. I >> suspect it's more a matter of "well, I didn't touch what broke, so >> it's not my priority." It would be nice to get back to our older team >> culture of "when the tree breaks, it's everyone's problem until it's >> fixed", with the sheriff being a backstop. I'm not sure how to do >> that except to keep beating the drum for a while. >> >> --Amanda >> > -- "Portability is generally the result of advance planning rather than trench warfare involving #ifdef" -- Henry Spencer (1992) --~--~---------~--~----~------------~-------~--~----~ Chromium Developers mailing list: [email protected] View archives, change email options, or unsubscribe: http://groups.google.com/group/chromium-dev -~----------~----~----~----~------~----~------~--~---
