Closed
Bug 1358270
Opened 8 years ago
Closed 8 years ago
Investigate an increase in the release crash ping size
Categories
(Data Platform and Tools :: General, enhancement)
Data Platform and Tools
General
Tracking
(Not tracked)
RESOLVED
FIXED
People
(Reporter: trink, Unassigned)
Details
The average message size is growing about 5% an hour starting around 2017-04-20T03:00:00Z
Reporter | ||
Comment 1•8 years ago
|
||
I assume the above is due to a change that has not been communicated out. However, the bigger problem is the raw number of crash pings has increased by over 20% in the same time period which tracks with the throttling of the new release.
Comment 2•8 years ago
|
||
I believe the size increase is due to the a recent change in the client that adds an additional payload (stacktraces) to the crash pings. :ddurst can probably confirm that.
Updated•8 years ago
|
Flags: needinfo?(ddurst)
Comment 3•8 years ago
|
||
Yes, both of those things should be expected as 53 goes from beta to release (and should be consistent if we compared size and rate to beta 52). If I should email a wider distro about this, I'm happy to do it. Who should I send to?
Flags: needinfo?(ddurst) → needinfo?(mtrinkala)
Updated•8 years ago
|
Component: Metrics: Pipeline → Pipeline Ingestion
Product: Cloud Services → Data Platform and Tools
Reporter | ||
Comment 4•8 years ago
|
||
fhr-dev
FYI the crash ping submission rate is still rising week over week and my personal experience with release 53 has been less than stellar in terms of crashes and hangs.
Flags: needinfo?(mtrinkala)
Comment 5•8 years ago
|
||
fhr-dev was notified on April 24.
But let's talk about this in some more detail:
a. "crash ping submission rate is still rising week over week" -- can I see this somewhere? Are you only talking about release?
b. "release 53 has been less than stellar in terms of crashes and hangs" -- is this related to the pings themselves, or just an observation about the performance of the release in general?
Flags: needinfo?(mtrinkala)
Reporter | ||
Comment 6•8 years ago
|
||
a) https://pipeline-cep.prod.mozaws.net/dashboard_output/graphs/analysis.moz_telemetry_doctype_monitor_crash.volume.html (it basically doubled the week before and this week it is up another ~50%). Yes, it may be leveling out with user adoption but having to wait weeks to know if there is a real problem or if it is the change to the submission policy is sub optimal.
b) observation running Release 53 on a Windows 10 laptop
Flags: needinfo?(mtrinkala)
Comment 7•8 years ago
|
||
closing as investigation is done.
Status: NEW → RESOLVED
Closed: 8 years ago
Resolution: --- → FIXED
Assignee | ||
Updated•3 years ago
|
Component: Pipeline Ingestion → General
You need to log in
before you can comment on or make changes to this bug.
Description
•