Bug 60579 - EWS bots should be robust against test-webkitpy hangs
Summary: EWS bots should be robust against test-webkitpy hangs
Status: RESOLVED FIXED
Alias: None
Product: WebKit
Classification: Unclassified
Component: New Bugs (show other bugs)
Version: 528+ (Nightly build)
Hardware: Unspecified Unspecified
: P2 Normal
Assignee: Eric Seidel (no email)
URL:
Keywords:
Depends on:
Blocks: 57724
  Show dependency tree / graph
 
Reported: 2011-05-10 14:03 PDT by Eric Seidel (no email)
Modified: 2011-05-10 22:32 PDT (History)
4 users (show)

See Also:

Attachments
Patch (3.15 KB, patch)
2011-05-10 14:04 PDT, Eric Seidel (no email) 		no flags Details | Formatted Diff | Diff
updated changelog (3.83 KB, patch)
2011-05-10 14:09 PDT, Eric Seidel (no email) 		no flags Details | Formatted Diff | Diff
now ignoring chromium gclient content too (3.99 KB, patch)
2011-05-10 14:18 PDT, Eric Seidel (no email) 		no flags Details | Formatted Diff | Diff
Patch (3.83 KB, patch)
2011-05-10 17:16 PDT, Eric Seidel (no email) 		abarth: review+
abarth: commit-queue-
Details | Formatted Diff | Diff
Show Obsolete (3) View All Add an attachment (proposed patch, testcase, etc.)

Note You need to log in before you can comment on or make changes to this bug.
Description Eric Seidel (no email) 2011-05-10 14:03:59 PDT 
EWS bots should be robust against test-webkitpy hangs
Comment 1 Eric Seidel (no email) 2011-05-10 14:04:35 PDT 
Created attachment 93003 [details]
Patch
Comment 2 Eric Seidel (no email) 2011-05-10 14:09:44 PDT 
Created attachment 93005 [details]
updated changelog
Comment 3 Eric Seidel (no email) 2011-05-10 14:13:28 PDT 
The only problem with this patch as written is that it will remove all the gclient directories for Chromium.  I'm not sure if that matters or not.  It might slow down the chromium EWSes more than we want.  I could add another --exclude line which ignores Source/WebKit/chromium
Comment 4 Adam Barth 2011-05-10 14:15:06 PDT 
Comment on attachment 93005 [details]
updated changelog

View in context: https://bugs.webkit.org/attachment.cgi?id=93005&action=review

> Tools/EWSTools/start-queue.sh:61
> +  git clean --force -x -d --exclude=WebKitBuild

I used to try this, but it slams the EBS volumes too much.
Comment 5 Eric Seidel (no email) 2011-05-10 14:16:40 PDT 
(In reply to comment #4)
> (From update of attachment 93005 [details])
> View in context: https://bugs.webkit.org/attachment.cgi?id=93005&action=review
> 
> > Tools/EWSTools/start-queue.sh:61
> > +  git clean --force -x -d --exclude=WebKitBuild
> 
> I used to try this, but it slams the EBS volumes too much.

What do you mean by that?
Comment 6 Eric Seidel (no email) 2011-05-10 14:18:36 PDT 
Created attachment 93007 [details]
now ignoring chromium gclient content too
Comment 7 Tony Chang 2011-05-10 14:18:51 PDT 
(In reply to comment #3)
> The only problem with this patch as written is that it will remove all the gclient directories for Chromium.  I'm not sure if that matters or not.  It might slow down the chromium EWSes more than we want.  I could add another --exclude line which ignores Source/WebKit/chromium

Doesn't git clean ignore the directories in .gitignore, which includes the gclient pulled directories?
Comment 8 Eric Seidel (no email) 2011-05-10 14:19:24 PDT 
git clean does, unless you pass -x. :)
Comment 9 Adam Barth 2011-05-10 14:30:30 PDT 
Comment on attachment 93007 [details]
now ignoring chromium gclient content too

View in context: https://bugs.webkit.org/attachment.cgi?id=93007&action=review

> Tools/EWSTools/start-queue.sh:63
> +  git clean --force -x -d --exclude=WebKitBuild --exclude=Source/WebKit/chromium

We should just use

find . -name "*.pyc" | xargs rm

to start with.  If we need to do mores we can revise.  -f -x -d is too big a hammer.
Comment 10 Eric Seidel (no email) 2011-05-10 16:45:19 PDT 
I think I tried to do too much in one patch.  I'll post a patch which has no functional changes besides the watchdog.
Comment 11 Eric Seidel (no email) 2011-05-10 17:16:44 PDT 
Created attachment 93053 [details]
Patch
Comment 12 Adam Barth 2011-05-10 18:38:14 PDT 
Looks fine, but the changelog isn't right anymore.
Comment 13 Eric Seidel (no email) 2011-05-10 22:32:23 PDT 
Committed r86213: <http://trac.webkit.org/changeset/86213>