595292 - browser_394759.js seems to hit network with dns failures

Status: RESOLVED FIXED

(Firefox :: Session Restore, defect)

Description

[Paul O'Shannessy [:zpao] (not bugmail, email directly)]

At least when running the test locally. It doesn't cause failures, but we shouldn't do it.

It's because we try to go to windowN.example.com. Simon actually brought this up in review, and I said meh because it didn't actually matter.

But now it's annoying to see comcast's dns error page in tests, so I want to fix it.

Comment 1

[(no longer active)]

I bet this is actually the cause of bug 570192.  It might make the test time out because of DNS query timeout for example (in which case the load event won't fire, I think).  The fix is really simple, and I think it's worth taking.

Comment 2

[(no longer active)]

Attachment: Patch (v1)

Comment 3

[Paul O'Shannessy [:zpao] (not bugmail, email directly)]

Comment on attachment 541106 [details] [diff] [review]
Patch (v1)

Review of attachment 541106 [details] [diff] [review]:
-----------------------------------------------------------------

I've been meaning to go back through and find bugs like this. Thanks for just doing it.

Comment 4

[Mounir Lamouri (:mounir)]

Pushed:
http://hg.mozilla.org/mozilla-central/rev/b060d115f501

Comment 5

[Jeff Walden [:Waldo]]

(In reply to comment #0)
comcast++ (!)

Comment 6

[Matt Brubeck (:mbrubeck)]

Comment on attachment 541106 [details] [diff] [review]
Patch (v1)

Requesting mozilla-approval-aurora.  This is a one-line patch to prevent a test from requesting a remote page over the internet.  The test became permaorange starting yesterday because of some infrastructure change or external change, and is still permaorange on Aurora.  So far releng has not been able to figure out what might have changed to trigger the failures.  This patch has landed on mozilla-central so the test is no longer failing there.

Landing this test-only change on Aurora now will allow us to have green tests for Firefox 6 as it goes through Aurora, Beta, and Release.

Comment 7

[Johnny Stenback (:jst)]

Comment on attachment 541106 [details] [diff] [review]
Patch (v1)

a=drivers for this test-only change.

Comment 8

[Daniel Holbert [:dholbert]]

http://hg.mozilla.org/releases/mozilla-aurora/rev/3b5e85a40598

Comment 9

[Virgil Dicu [:virgil] [QA]]

Hello.

Is there a way in which this issue can be verified? Some simple STR or a test case?