1242727 - Expand BHR pseudo-stack coverage of e10s *child* process

Status: RESOLVED WORKSFORME

(Toolkit :: Telemetry, defect, P3)

Description

[Vladan Djeric (:vladan)]

In bug 1224374, Yoric added additional labels for BHR pseudostack frames based on chromehang data from http://telemetry.mozilla.org/chromehangs

However, chrome-hang reporting is parent-process only on e10s, and e10s is the active configuration on 86% of Nightly (where the data comes from).

We need to do an analysis of top chromehang stacks from single-process Nightly profiles only. This would tell us where we are lacking BHR pseudostack coverage of content code.

Comment 1

[Vladan Djeric (:vladan)]

Roberto & poiru: can one of you guys do this analysis of top single-process chromehangs? Alternately, you could add an e10s selector to the chromehangs dashboard instead

Yoric: can you add the missing pseudostack labels as a result of their analysis

Comment 2

[Roberto Agostino Vitillo (:rvitillo)]

Assigning to poiru as he is working on Bug 1240887.

Comment 3

[David Teller [:Yoric] - still alive but not very active]

So, waiting for poiru.

Comment 4

[David Teller [:Yoric] - still alive but not very active]

Well, apparently, this will need a custom analysis.
I'll try and get to this on Monday.
For reference, I'll need to read this: http://robertovitillo.com/2015/01/16/next-gen-data-analysis-framework-for-telemetry/

Comment 5

[David Teller [:Yoric] - still alive but not very active]

Actually, azhang jumped in and got me the data. Thanks, azhang!

So now, it's about adding the BHR labels for the following list:
https://gist.github.com/Uberi/9348e5945d8434e1068e

Comment 6

[David Teller [:Yoric] - still alive but not very active]

Also, https://gist.github.com/Uberi/9348e5945d8434e1068e

Comment 7

[David Teller [:Yoric] - still alive but not very active]

Well, this turns out to be very simple. All the items of the list already have BHR labels.