Closed Bug 988015 Opened 7 years ago Closed 7 years ago

Disable seer on B2G

Categories

(Core :: Networking, defect)

ARM
Gonk (Firefox OS)
defect
Not set
normal

Tracking

()

RESOLVED FIXED
mozilla31
Tracking Status
firefox29 --- unaffected
firefox30 --- fixed
firefox31 --- fixed
b2g-v1.3 --- fixed
b2g-v1.3T --- fixed
b2g-v1.4 --- fixed
b2g-v2.0 --- fixed

People

(Reporter: u408661, Assigned: u408661)

Details

Attachments

(1 file)

From discussion with :jduell, we're not sure it even works right on B2G due to limited file I/O in child processes, and the memory overhead of an sqlite connection per child is just too great. We'll turn it back on once we either have the seer running entirely in the parent or have a rewritten backend.
Attached patch patchSplinter Review
try push: https://tbpl.mozilla.org/?tree=Try&rev=86d4904e1f62
Attachment #8396822 - Flags: review?(jduell.mcbugs)
Attachment #8396822 - Flags: review?(jduell.mcbugs) → review+
Is this something we'll need on branches?
(In reply to Kyle Huey [:khuey] (khuey@mozilla.com) from comment #2)
> Is this something we'll need on branches?

Unsure... what is the most recent non-trunk branch that exists? (My lack of B2G hacking is showing)
We're going to want this for all branches since the seer landed--aka bug 881804, right Nick?
Flags: needinfo?(hurley)
(In reply to Jason Duell (:jduell) from comment #5)
> We're going to want this for all branches since the seer landed--aka bug
> 881804, right Nick?

Sounds about right, though I don't know which of those branches affect b2g... b2g versioning, it is a mystery (to me).
Flags: needinfo?(hurley)
https://hg.mozilla.org/mozilla-central/rev/e6eee5dd8c73
Status: NEW → RESOLVED
Closed: 7 years ago
Resolution: --- → FIXED
Target Milestone: --- → mozilla31
Well if it landed for 27 then we need this on Aurora (which is b2g 1.4) and b2g-28 (which is b2g 1.3).
Comment on attachment 8396822 [details] [diff] [review]
patch

Thanks Kyle.

[Approval Request Comment]
Bug caused by (feature/regressing bug #):  bug 881804
User impact if declined: we'll waste disk space and RAM on B2G without getting any perf win.
Risk to taking this patch (and alternatives if risky): very low--just turns new feature off.
String or UUID changes made by this patch: no

> NOTE: B2G-18 is now open for security fixes only.

This is just a pref-off of some code that's just wasting disk and RAM space, so I'd still recommend it ride along on b2g-18 if possible.
Attachment #8396822 - Flags: approval-mozilla-b2g28?
Attachment #8396822 - Flags: approval-mozilla-b2g18?
Attachment #8396822 - Flags: approval-mozilla-aurora?
Attachment #8396822 - Flags: approval-mozilla-b2g18?
Attachment #8396822 - Flags: approval-mozilla-aurora? → approval-mozilla-aurora+
(In reply to Jason Duell (:jduell) from comment #9)
> Comment on attachment 8396822 [details] [diff] [review]
> patch
> 
> Thanks Kyle.
> 
> [Approval Request Comment]
> Bug caused by (feature/regressing bug #):  bug 881804
> User impact if declined: we'll waste disk space and RAM on B2G without
> getting any perf win.
> Risk to taking this patch (and alternatives if risky): very low--just turns
> new feature off.
> String or UUID changes made by this patch: no
> 
> > NOTE: B2G-18 is now open for security fixes only.
> 
> This is just a pref-off of some code that's just wasting disk and RAM space,
> so I'd still recommend it ride along on b2g-18 if possible.

We are landing only critical blocker bugs on 1.3(b2g 28) branch right now. This bug would be an exception given, we want to turn-off a feature that's not ready, I am ok to make an exception and take this.

Is this setting being used anywhere to worry about fallouts here ?
Flags: needinfo?(jduell.mcbugs)
> Is this setting being used anywhere to worry about fallouts here ?

Should be no fallout.  Just turning off some predictor optimizations that don't actually work but consume CPU/IO.  Very low risk.

Nick, please make sure these get landed on the appropriate trees (I forget if there's a particular flag we can set to make that happen: otherwise hg push 'em).  Wait for :bajaj's approval first of course :)
Flags: needinfo?(jduell.mcbugs)
Attachment #8396822 - Flags: approval-mozilla-b2g28? → approval-mozilla-b2g28+
This bug is really fixed as-filed. Setting the status to disabled really means we disabled the functionality added by this bug on the branch.
You need to log in before you can comment on or make changes to this bug.