Bug 170678 - webkitpy: Increase estimate for processes run by iOS Simulators
Summary: webkitpy: Increase estimate for processes run by iOS Simulators
Status: RESOLVED FIXED
Alias: None
Product: WebKit
Classification: Unclassified
Component: Tools / Tests (show other bugs)
Version: WebKit Nightly Build
Hardware: Unspecified Unspecified
: P2 Normal
Assignee: Jonathan Bedard
URL:
Keywords: InRadar
Depends on:
Blocks:
 
Reported: 2017-04-10 09:38 PDT by Jonathan Bedard
Modified: 2017-04-11 08:22 PDT (History)
7 users (show)

See Also:

Attachments
Patch (1.53 KB, patch)
2017-04-10 09:43 PDT, Jonathan Bedard 		no flags Details | Formatted Diff | Diff
Patch for landing (1.53 KB, patch)
2017-04-10 10:43 PDT, Jonathan Bedard 		no flags Details | Formatted Diff | Diff
Show Obsolete (1) 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 Jonathan Bedard 2017-04-10 09:38:39 PDT 
Previously, it was estimated that each simulated iOS device would run 100 processes.  With newer version of iOS, this is closer to 125.
Comment 1 Radar WebKit Bug Importer 2017-04-10 09:39:55 PDT 
<rdar://problem/31534061>
Comment 2 Jonathan Bedard 2017-04-10 09:43:46 PDT 
Created attachment 306693 [details]
Patch
Comment 3 Aakash Jain 2017-04-10 09:53:20 PDT 
Comment on attachment 306693 [details]
Patch

looks good to me.
Comment 4 Jonathan Bedard 2017-04-10 10:00:04 PDT 
Will commit queue once EWS comes back clean.
Comment 5 Jonathan Bedard 2017-04-10 10:43:13 PDT 
Created attachment 306708 [details]
Patch for landing
Comment 6 WebKit Commit Bot 2017-04-10 11:12:00 PDT 
Comment on attachment 306708 [details]
Patch for landing

Clearing flags on attachment: 306708

Committed r215191: <http://trac.webkit.org/changeset/215191>
Comment 7 WebKit Commit Bot 2017-04-10 11:12:04 PDT 
All reviewed patches have been landed.  Closing bug.
Comment 8 Alexey Proskuryakov 2017-04-10 21:13:50 PDT 
Will we also change the plist on the wiki to allow for more files?
Comment 9 Jonathan Bedard 2017-04-11 08:19:33 PDT 
Talking with Aakash yesterday, this is likely not the only change we need.  We should consider re-doing the tests he did to determine the optimal number of simulators to run.  He sent me his measurements from awhile back, on newer simulators, I suspect somewhere around 8 is optimal, as opposed to 12.

We haven't seen the file cap be an issue.
Comment 10 Aakash Jain 2017-04-11 08:22:09 PDT 
(In reply to Alexey Proskuryakov from comment #8)
> Will we also change the plist on the wiki to allow for more files?

both the plist files are still good, since they increase the process, file limits significantly. The change in this bug is very minor.