[Webkit-unassigned] [Bug 63610] [NRWT] make nrwt able to test the same order like orwt

bugzilla-daemon at webkit.org bugzilla-daemon at webkit.org
Thu Jul 7 19:29:36 PDT 2011


Dirk Pranke <dpranke at chromium.org> changed:

           What    |Removed                     |Added
  Attachment #99971|commit-queue?               |
               Flag|                            |

--- Comment #12 from Dirk Pranke <dpranke at chromium.org>  2011-07-07 19:29:35 PST ---
(From update of attachment 99971)
Okay, I think I get it now. This patch is trying to do two things ...

1) Move the HTTP and websocket tests to the end

2) Create shards of tests, 1000 at a time.

It is true that ORWT does (1). 

I think (2) is wholly new, correct? ORWT didn't have a "sharding" concept, and sharding simply by a fixed number probably won't work all that well with intra-directory dependencies (you really want shard-by-directory instead, or possibly fully_parallel if you don't really care about how things are sharded). Note that the sharding is different from just running subsets of the tests. There are other flags that control that, prior to creating the shards.

In addition, one problem with moving the http tests to the end is that there no good way to coordinate this with the http locking code that we use to ensure that we don't have two NRWT processes running at once that are trying to start and stop web servers. Another problem is that the HTTP tests are the slowest tests, so you really want to start them right away when running multiple processes, to minimize the overall delay time. 

See bug 58625 for a fuller discussion of this. It's possible that we can fix the coordination of locking and unlocking and still put the http tests at the end, but that work will be nontrivial and, like I said, I don't think you want to put the tests at the end with multiple processes anyway (so if anything, I would suggest you add a patch that puts the http tests at the end only if in single-process mode, and decide that you don't care or need the concurrency).

If after all I've written you don't think NRWT will do what you want, maybe you can describe what the test conditions you are trying to replicate are and we can see what might make the most sense?

Configure bugmail: https://bugs.webkit.org/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are the assignee for the bug.

More information about the webkit-unassigned mailing list