-
Notifications
You must be signed in to change notification settings - Fork 26.7k
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
[CI/FTL] Oriole to Panther, presubmit false #130912
Conversation
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Did you mean to include these files? If so can you add something to the description that explains why.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
I did not
.ci.yaml
Outdated
@@ -411,8 +411,9 @@ targets: | |||
- bin/** | |||
- .ci.yaml | |||
|
|||
- name: Linux firebase_oriol33_abstract_method_smoke_test | |||
- name: Linux firebase_panther33_abstract_method_smoke_test |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Can we skip the device model info in the name?
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
It was originally split out like this to avoid conflict with another similarly named test that runs on a number of devices.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
(by @godofredoc )
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
I would expect we will have other model change in the future, and each change will cut off the historical builds. Does it make sense to use a more general name for these two firebase tasks?
These are what we have (or proposed in this PR)
Linux firebase_abstract_method_smoke_test
(based on griffin │ XT1650 | os:24
and redfin │ Pixel 5 | os:30
)
Linux firebase_panther33_abstract_method_smoke_test
(panther | pixel 7 | os: 33
)
Does it make sense to use low_end v.s. high_end, or something similar instead?
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Maybe we could just merge this back in with the abstract_method_smoke_test
because panther shoul dbe more available than oriole, and the problem was that oriole was causing timeouts.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Based on #130499 (comment), the capacity of griffin
is marked as low
. We may hit similar issues for this model in the future.
Not sure about the main differences between these three devices. Is it the os that we care and want to test against? If yes, I would separate/rename the target based on their os_version instead of the model (assuming the os doesn't change and the model changes).
If we are confident the capacity is no longer an issue any more, it sgtm to merge them. But we need to find an replacement of the griffin with a device model of high/medium capacity. otherwise we will hit the same issue again.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
The goal is to test on some highly available physical devices and make sure we're targetting as many API levels as possible. But there's no automated monitoring right now (beyond flakiness) about whether a device has become lower in availablility or not.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
In that case, can we replace griffin
with a high available device model with the same or close API level? Considering panther
is high now, it makes to me to consolidate it back to abstract_method_smoke_test
.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Ok, I moved panther to the old one, removed griffin and instead test API 24 on a virtual device, and mark them all as presubmit: false so we don't overload FTL.
.ci.yaml
Outdated
@@ -444,8 +427,8 @@ targets: | |||
task_name: abstract_method_smoke_test | |||
physical_devices: >- | |||
[ | |||
"--device", "model=panther,version=33", | |||
"--device", "model=redfin,version=30", |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
No comma is expected for the last element in the property list, and this is causing the test ownership task failure.
(This PR is still not merging despite having autosubmit and approvals) |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
LGTM
F.Y.I. @ricardoamador Could you help take a look at this PR? seems the bot misses it. Is it related to some cocoon version switch around that time? I just removed and re-added the label. |
@keyonghan taking a look now. |
Wait, this PR is being submitted to |
Yeah I just saw that. I don't think it should matter as it would be treated as we treat the release branches. |
Hah, whups |
Maybe the bot should have warned me not to try to merge into main? |
The bot could provide a warning but we allow this for the purposes of the release team. It looks like something is providing an ignore temporarily status but I have not been able to find it. |
I would be surprised if this merged after changing target to master. But if it does that will help give me a clue to where we were hung. |
Hmm, are there any branch protections on main? For some reason when the target branch was changed to master I had the ability to squash and merge but with main I did not. I am wondering if this was the same for the bot? Okay this might merge then. Which will tell us this must be a branch protection issue. |
Main appears to have branch protections. I could not manually merge it there, but the button is enabled now that I switched it to master. |
Looks like there were some protections in place with main that did not allow the merge. Though I think the bot should have at least logged this. I am going to open an issue to see if we can detect any of this. |
Created two issues as a result of Autosubmit refusing to do its job. |
flutter/flutter@9cfbf6b...e8b397c 2023-07-22 polinach@google.com Setup leak tracking regression for material. (flutter/flutter#130169) 2023-07-22 engine-flutter-autoroll@skia.org Roll Flutter Engine from 6344b17a2e03 to 481684a6e276 (2 revisions) (flutter/flutter#131118) 2023-07-22 engine-flutter-autoroll@skia.org Roll Flutter Engine from b47cf14fda0e to 6344b17a2e03 (1 revision) (flutter/flutter#131114) 2023-07-22 engine-flutter-autoroll@skia.org Roll Flutter Engine from 840bcc3449ff to b47cf14fda0e (3 revisions) (flutter/flutter#131109) 2023-07-22 engine-flutter-autoroll@skia.org Roll Flutter Engine from 2d8cff44261b to 840bcc3449ff (11 revisions) (flutter/flutter#131101) 2023-07-21 goderbauer@google.com Remove obsolete work around for shadow drawing (flutter/flutter#131066) 2023-07-21 engine-flutter-autoroll@skia.org Manual roll Flutter Engine from acb5d0640b6c to 2d8cff44261b (3 revisions) (flutter/flutter#131092) 2023-07-21 polinach@google.com Upgrade to newer leak_tracker. (flutter/flutter#131085) 2023-07-21 engine-flutter-autoroll@skia.org Manual roll Flutter Engine from f5c1650c7acc to acb5d0640b6c (10 revisions) (flutter/flutter#131070) 2023-07-21 32538273+ValentinVignal@users.noreply.github.com Suggest a potential valid name for the flutter project when using `flutter create` (flutter/flutter#130900) 2023-07-21 dnfield@google.com [CI/FTL] Oriole to Panther, presubmit false (flutter/flutter#130912) 2023-07-21 6655696+guidezpl@users.noreply.github.com Improve handling of certain icons in RTL (flutter/flutter#130979) If this roll has caused a breakage, revert this CL and stop the roller using the controls here: https://autoroll.skia.org/r/flutter-packages Please CC rmistry@google.com,stuartmorgan@google.com on the revert to ensure that a human is aware of the problem. To file a bug in Packages: https://github.com/flutter/flutter/issues/new/choose To report a problem with the AutoRoller itself, please file a bug: https://bugs.chromium.org/p/skia/issues/entry?template=Autoroller+Bug Documentation for the AutoRoller is here: https://skia.googlesource.com/buildbot/+doc/main/autoroll/README.md
Alternative to flutter#130868 Oriole is the Pixel 6, Panther is the Pixel 7. Panther is more available in FTL at this point. There's less value in running this on presubmit, since those can spawn many jobs multiple times over as people push new commits to branches. Let's just run it post submit to avoid overloading the capacity of FTL.
Alternative to flutter#130868 Oriole is the Pixel 6, Panther is the Pixel 7. Panther is more available in FTL at this point. There's less value in running this on presubmit, since those can spawn many jobs multiple times over as people push new commits to branches. Let's just run it post submit to avoid overloading the capacity of FTL.
Alternative to #130868
Oriole is the Pixel 6, Panther is the Pixel 7. Panther is more available in FTL at this point.
There's less value in running this on presubmit, since those can spawn many jobs multiple times over as people push new commits to branches. Let's just run it post submit to avoid overloading the capacity of FTL.