Rename raptor-tp6 subtests to remove redundant test name prefix
Categories
(Testing :: Raptor, enhancement, P1)
Tracking
(firefox66 fixed)
Tracking | Status | |
---|---|---|
firefox66 | --- | fixed |
People
(Reporter: davehunt, Assigned: igoldan)
References
Details
Attachments
(2 files)
The subtest names for raptor-tp6 repeat the test name, contributing to issues such as bug 1517512. We should clean these up, so tests such as:
- raptor-tp6-amazon-firefox-dcf
- raptor-tp6-amazon-firefox-fnbpaint
- raptor-tp6-amazon-firefox-ttfi
- raptor-tp6-amazon-firefox-hero:hero1
- raptor-tp6-amazon-firefox-loadtime
Are renamed to:
- dcf
- fnbpaint
- ttfi
- hero:hero1
- loadtime
Reporter | ||
Comment 1•5 years ago
|
||
We're not currently alerting on these, so we should be safe for us to test our solution to bug 1520170 when we rename these.
Assignee | ||
Updated•5 years ago
|
Assignee | ||
Updated•5 years ago
|
Assignee | ||
Comment 2•5 years ago
|
||
On 1st look, this doesn't look like a basic *.ini tweak. The following line from output.py generates the long version of the name:
new_subtest['name'] = test.name + "-" + measurement_name
Updating it would change names for all our Raptor tp6 subtests. Benchmarks will remain the same, but they're already ok.
Assignee | ||
Comment 3•5 years ago
|
||
Comment 5•5 years ago
|
||
(In reply to Ionuț Goldan [:igoldan], Performance Sheriffing from comment #4)
:rwood could you review this PR, please?
Yes I'll have a look, thanks. Do you have a try push? Also are there any implications for this in perfherder - you know more about perfherder than I do, so I'm assuming you've already covered that. :)
Comment 6•5 years ago
|
||
(In reply to Robert Wood [:rwood] from comment #5)
Yes I'll have a look, thanks. Do you have a try push? Also are there any implications for this in perfherder - you know more about perfherder than I do, so I'm assuming you've already covered that. :)
Ah - I found the try link you noted in phabricator, thanks
https://treeherder.mozilla.org/#/jobs?repo=try&revision=b3dcb6495f64d1790cb9b4722ca36cfa7ac53c86
Assignee | ||
Comment 7•5 years ago
|
||
(In reply to Robert Wood [:rwood] from comment #6)
Also are there any implications for this in perfherder - you know more about perfherder than I do, so I'm assuming you've already covered that. :)
It shouldn't cause problems. I double checked the initial ingestion phase and it looks ok.
On the FE side, the Javascript shouldn't have problems with the shorter name form; it should treat it as it is, without looking for any patterns. I already found recent Try signatures with dcf/fcp/etc subtests in the Graphs view, backing that up.
Pushed by igoldan@mozilla.com: https://hg.mozilla.org/integration/autoland/rev/9c24a1d10557 Do not combine test name with measurement r=rwood
Comment 9•5 years ago
|
||
bugherder |
Assignee | ||
Updated•5 years ago
|
Assignee | ||
Comment 10•5 years ago
|
||
Reopening as this hasn't been handled on Perfherder yet.
Updated•5 years ago
|
Assignee | ||
Updated•5 years ago
|
Comment 11•5 years ago
|
||
Reporter | ||
Updated•5 years ago
|
Assignee | ||
Comment 12•5 years ago
|
||
Joel, could you have a look over the PR from comment 11? I need someone's review, before merging & running the management script on production.
Assignee | ||
Comment 14•5 years ago
•
|
||
(In reply to Joel Maher ( :jmaher ) (UTC-4) from comment #13)
nice PR- the logic looks sound
Thanks! I've just sent you the review request.
Comment 15•5 years ago
|
||
bugherder uplift |
Updated•5 years ago
|
Assignee | ||
Comment 16•5 years ago
|
||
The new script got deployed on production. It ran successfully. Thus, we can close this bug.
Description
•