Limit recording engagement scalars to parent process




a month ago
a month ago


(Reporter: gfritzsche, Unassigned, Mentored)


(Blocks: 1 bug)


Firefox Tracking Flags

(firefox59 affected)


(Whiteboard: [good first bug] [lang=js])



a month ago
Bug 1423446 shows that some engagement scalars are not recorded in the content process:

We should restrict these scalars to record only in the parent, i.e.:
> record_in_processes: ["main"]

This change is needed in Scalars.yaml:

After the change, make sure that building still works properly using:
> mach build

Also run some tests to make sure nothing broke:
> mach test browser/modules/test/browser/browser_UsageTelemetry*


a month ago
Blocks: 1423446

Comment 1

a month ago
While we're here, we can also restrict these two:

Comment 2

a month ago

I just started with mozilla development and would like to fix this.Just need a few resousrce to know how to fix it?
Can anyone help me out here?

Hi Rohan!

As outlined in comment 0, the "record_in_processes" property needs to be set to "main" only for the scalars listed in comment 0 and 1.
Assignee: nobody → rohan17089

Comment 4

a month ago
Thanks for the quick reply!

I did check all these scalars and it is already set to main only.
Couldn't find the following scalar though
perhaps telementry might right to another type of build
and for the following scalar
the main written wasnt in single quotes,don't know if it makes a difference or not

This is all what i could figure out,waiting for your further reply.
(In reply to Rohan Rajpal from comment #4)
> Thanks for the quick reply!
> I did check all these scalars and it is already set to main only.

Good point. I cross-checked and you're right, all these scalars are already set to "main". Rohan, I think you're done here :) I've needinfo'd you on another good bug if you're interested in taking more!

@Frank, looks like the list of scalars from comment 0 comes from bug 1423446: we shouldn't be seeing these scalars in the content processes though. I quickly checked my builds and looks like we're sending the right stuff. Is this a mistake in the list or some unexpected behaviour in the pipeline code for scalars?
Assignee: rohan17089 → nobody
Flags: needinfo?(fbertsch)
Everything looks correct:
- The list I gave is all probes that have *only* reported in the parent process
- This bug is for restricting scalars to those
- It looks like those scalars are already restricted, so we're good to go!
Flags: needinfo?(fbertsch)
Thanks Frank!
Last Resolved: a month ago
Resolution: --- → WORKSFORME
You need to log in before you can comment on or make changes to this bug.