Closed Bug 1366394 Opened 5 years ago Closed 5 years ago
More than 3% of Safe Browsing updates time out
59 bytes, text/x-review-board-request
According to the telemetry we added in bug 1336904, more than 3% of users are timing out while trying to contact the Safe Browsing update servers (both Google and Mozilla): https://telemetry.mozilla.org/new-pipeline/dist.html#!cumulative=0&end_date=2017-05-18&keys=google4!google!mozilla!mozilla&max_channel_version=nightly%252F55&measure=URLCLASSIFIER_UPDATE_TIMEOUT&min_channel_version=null&processType=*&product=Firefox&sanitize=1&sort_keys=submissions&start_date=2017-05-03&table=1&trim=1&use_submission_date=0 We should bump the timeout since this is too high.
Comment on attachment 8869596 [details] Bug 1366394 - Bump the Safe Browsing update timeout to 15s. https://reviewboard.mozilla.org/r/141176/#review144954 Thanks Francois :) Just out of curiosity, how do we come out with the new timeout value? Does it come from another telemetry probe? (We seemed to discuss about this issue but cannot clear recall) Thanks!
Attachment #8869596 - Flags: review?(hchang) → review+
Pushed by firstname.lastname@example.org: https://hg.mozilla.org/integration/autoland/rev/9353b09c2ec9 Bump the Safe Browsing update timeout to 15s. r=hchang
Thanks for the review Henry. (In reply to Henry Chang [:hchang] from comment #3) > Just out of curiosity, how do we come out with the new timeout value? > Does it come from another telemetry probe? Here's my reasoning for the 15 seconds: I made it 3 times larger in the hope that this we will see 3 times less timeouts. It seemed like a large enough number to allow very networks to get updates, but it's also 4 times smaller than the overall update timeout (60 seconds). I plan on monitoring the timeouts with this new value and re-adjusting again if necessary.
You need to log in before you can comment on or make changes to this bug.