Enforce disk quota on pending pings
Categories
(Data Platform and Tools :: Glean: SDK, enhancement, P1)
Tracking
(Not tracked)
People
(Reporter: Dexter, Assigned: brizental)
References
Details
(Whiteboard: [telemetry:glean-rs:m16])
Attachments
(2 files, 1 obsolete file)
42 bytes,
text/x-github-pull-request
|
Details | Review | |
3.01 KB,
text/plain
|
chutten
:
data-review+
|
Details |
Glean does not currently check how much space is taken by pending pings, but we should, and should also have telemetry about how many times quota is hit. We should probably copy Firefox desktop mechanism as a starting point.
Reporter | ||
Updated•5 years ago
|
Reporter | ||
Updated•5 years ago
|
Reporter | ||
Updated•5 years ago
|
Assignee | ||
Updated•5 years ago
|
Assignee | ||
Comment 1•4 years ago
|
||
The Firefox mechanism not only has metrics to record the space taken by the pending pings directory, but also the time it took to calculate this space. I don't believe we need this time metric just yet for Glean, what do you think [:Dexter]?
I believe we could have only the metrics for
- Size of the pending pings directory;
- How many ping files were deleted due to extrapolating the size.
Now, the current Firefox mechanism records these numbers even if the quota has not been hit. I am not sure if that would be ideal for Glean too or if it should only record when the quota is hit.
Reporter | ||
Comment 2•4 years ago
|
||
(In reply to Beatriz Rizental from comment #1)
The Firefox mechanism not only has metrics to record the space taken by the pending pings directory, but also the time it took to calculate this space. I don't believe we need this time metric just yet for Glean, what do you think [:Dexter]?
I agree, I don't think we need that time.
I believe we could have only the metrics for
- Size of the pending pings directory;
- How many ping files were deleted due to extrapolating the size.
These sounds about right.
Now, the current Firefox mechanism records these numbers even if the quota has not been hit.
I am not sure if that would be ideal for Glean too or if it should only record when the quota is hit.
I believe the same rationale makes sense for Glean. Ideally, most of the scans happen when quota is not hit (i.e. when glean starts).
Assignee | ||
Comment 3•4 years ago
|
||
Comment 4•4 years ago
|
||
Assignee | ||
Updated•4 years ago
|
Comment 5•4 years ago
|
||
Assignee | ||
Comment 6•4 years ago
|
||
Comment 7•4 years ago
|
||
Assignee | ||
Updated•4 years ago
|
Description
•