Closed Bug 1813060 Opened 1 year ago Closed 6 months ago

Investigate performance impact of global CMP cookie banner rules


(Core :: Privacy: Anti-Tracking, task, P2)






(Reporter: pbz, Unassigned)



Supporting global rules means we need to run banner detection (and clicking) code on every site. We need to run performance tests to see if this is feasible. Open questions include:

  • Does our global rule mechanism regress page performance? If so, how much?
    • Test on Desktop
    • Test on Android
    • Test on slow devices
  • How many global rules can be deployed to minimize performance impact while still providing good banner handling coverage?

We can test by:

  • Running a set of global rules on Nightly to see if we get perf alerts or bugs filed
  • Run mach try perf with different amounts of global rules (e.g. browsertime)
  • Record performance profiles comparing feature fully disabled, site-specific rules enabled, global rules enabled
  • Adding performance markers for better analysis via the profiler
Blocks: cbh-cmp-perf
No longer blocks: cbh-cmp

We've investigated performance of the feature in collaboration with the perf team. Our investigation included feature specific perf telemetry and a Nimbus experiment in Nightly. There are no perf blockers for releasing the feature to a limited (release user) audience in private browsing. If we want to roll out further we need more investigation and probably more performance work.

Closed: 6 months ago
Resolution: --- → FIXED
You need to log in before you can comment on or make changes to this bug.