Investigate lib-crash invalid value errors in Fenix
Categories
(Data Platform and Tools :: Glean: SDK, task, P1)
Tracking
(Not tracked)
People
(Reporter: travis_, Assigned: travis_)
References
Details
(Whiteboard: [telemetry:glean-rs:m11])
Attachments
(1 file)
Assignee | ||
Updated•5 years ago
|
Assignee | ||
Comment 1•5 years ago
|
||
The way that this is implemented in lib-crash, it's counting all of the recorded errors in the persisted file and then calling add(<number of errors>)
, which could potentially be zero, and calling add(0)
is an invalid_value
error.
Assignee | ||
Comment 2•5 years ago
|
||
Assignee | ||
Updated•5 years ago
|
Assignee | ||
Comment 3•5 years ago
|
||
This is still happening at a low/lower frequency, reopening this to investigate again.
Assignee | ||
Updated•5 years ago
|
Assignee | ||
Comment 4•5 years ago
|
||
Looking at this more closely using our error dashboards, it doesn't appear that we have had an instance of this occur in Fenix nightly since the 19th, and have seen a sharp falloff of occurrences just before that. This is even better if filtering for versions of Fenix using Glean-core.
Even looking at this in FF-Preview Release, we are starting to see a downward trend in both counts and affected clients.
Based on Nightly showing little or no occurrences, and based on the trend in Release towards zero, I'm closing this as fixed.
Comment hidden (collapsed) |
Description
•