948569 - Disable seer on Beta 27 and Aurora 28

Status: RESOLVED FIXED

(Core :: Networking, defect)

Description

[u408661]

Attachment: disable_seer.patch

Until we have satisfactory solutions for bug 945779 and bug 947745 (harder than it looks), we should disable the seer on beta (and release, of course)

[Approval Request Comment]
Bug caused by (feature/regressing bug #): bug 881804
User impact if declined: possible high CPU usage during page load, unbounded sql disk usage
Testing completed (on m-c, etc.): ensured pref works as advertised
Risk to taking this patch (and alternatives if risky): low
String or IDL/UUID changes made by this patch: none

Comment 1

[u408661]

Comment on attachment 8345430 [details] [diff] [review]
disable_seer.patch

See comment 0 for approval justification.

Comment 2

[u408661]

We have other speculative connections than just those created by the seer :)

Comment 3

[Richard Newman [:rnewman]]

I landed this on Beta.

https://hg.mozilla.org/releases/mozilla-beta/rev/28c5128b8925

Comment 4

[Richard Newman [:rnewman]]

Looks like the tests are not expecting the feature to be disabled.

Comment 5

[Richard Newman [:rnewman]]

Landed a fix for the test:

https://hg.mozilla.org/releases/mozilla-beta/rev/ecfaa4a0cd8d

Comment 6

[Ryan VanderMeulen [:RyanVM]]

(In reply to Richard Newman [:rnewman] from comment #5)
> Landed a fix for the test:
> 
> https://hg.mozilla.org/releases/mozilla-beta/rev/ecfaa4a0cd8d

Please land this fix on trunk as well so we don't have to repeat this song and dance on every merge cycle.

Comment 7

[u408661]

Aurora push: remote:   https://hg.mozilla.org/releases/mozilla-aurora/rev/ed83e1080ed0

RyanVM: I'll land the test patch on m-c in a separate bug a bit later (though my plan is to have it be irrelevant for 29 anyway).

Comment 8

[u408661]

(In reply to Richard Newman [:rnewman] from comment #3)
> I landed this on Beta.

Thanks for this (and fixing my test-related bustage)!

Comment 9

[Ryan VanderMeulen [:RyanVM]]

(In reply to Nicholas Hurley [:hurley] from comment #7)
> RyanVM: I'll land the test patch on m-c in a separate bug a bit later
> (though my plan is to have it be irrelevant for 29 anyway).

Yes, but plans can change so it never hurts to be prepared :)