-
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
Mac_ios hot_mode_dev_cycle_ios__benchmark is 2.00% flaky, Process exited with status = -1 (0xffffffff) lost connection
#120808
Comments
The flake in 4015 is a host-side Dart VM GC crasher. I've only seen that once, and without a reliable repro, I don't think the VM team will be able to track that one down. However, 3995 is a VM service address discovery timeout. It is interesting because the device logs don't capture the timeframe when the discovery timed out. Instead the device logs seem to begin only slightly before the run that follows the failing run. cc @jmagman. Downgrading to P3 since this is only just over the threshold. |
[prod pool] current flaky ratio for the past (up to) 100 commits is 3.06%. Flaky number: 3; total number: 98. Recent test runs: |
Hm, 4094 and 4171 lost connection to device, not sure what that's about? It ran on different devicelab bots.
4160 is VM service discovery timeout. |
Routing to the infra ticket queue to investigate lost connection to the device. |
4201 and 4200 are Dart VM GC crashers on the host 4233 is an Observatory discovery timeout 4316, 4284, and 4274 are all instances of:
I wanted to exclude the possibility that this is caused by a real engine crash during startup, but unfortunately the device logs (like here for the run here) don't go back far enough to capture the failure. The GC crash is already flagged in a couple of other P1's, so we can those other issues to track it, as is the Observatory discovery timeout. Since we haven't excluded the possibility that this is an engine crash, I'm going to remove from the infra queue, and route to |
I re-opened #120809 to mark this test flaky. |
[prod pool] flaky ratio for the past (up to) 100 commits between 2023-08-10 and 2023-08-15 is 3.13%. Flaky number: 3; total number: 96. Recent test runs: |
[prod pool] flaky ratio for the past (up to) 100 commits between 2023-08-16 and 2023-08-22 is 5.26%. Flaky number: 5; total number: 95. Recent test runs: |
Each of the recent flakes are still the The idea is, if it errors when launching, try again (without throwing an error). It seems to always work on rerun so we can just bake retries into the launch code |
[prod pool] flaky ratio for the past (up to) 100 commits between 2023-08-23 and 2023-08-29 is 4.26%. Flaky number: 4; total number: 94. Recent test runs: |
Sometimes `ios-deploy` loses connection to the device after installing, starting debugserver, and launching. This is shown with an error message like: ``` Process 579 exited with status = -1 (0xffffffff) lost connection ``` This happens frequently in our CI system: #120808 Usually in CI, on retry it'll work and pass - so this is an attempt to retry without failing the test first. It's not guaranteed to fix since we're unable to recreate this error locally.
I pushed a workaround that should hopefully fix: #133769 Since we can't recreate it, though, we'll just have to wait to see if it stops flaking |
[prod pool] flaky ratio for the past (up to) 100 commits between 2023-08-30 and 2023-09-03 is 1.03%. Flaky number: 1; total number: 97. Recent test runs: |
[prod pool] flaky ratio for the past (up to) 100 commits between 2023-09-07 and 2023-09-12 is 1.03%. Flaky number: 1; total number: 97. Recent test runs: |
When retrying to connect to the device during app launch, don't uninstall the app first. Latest test flake for #120808: https://logs.chromium.org/logs/flutter/buildbucket/cr-buildbucket/8770202475999850785/+/u/run_hot_mode_dev_cycle_ios__benchmark/test_stdout Shows that it uninstalled and then tried debugging and failed, which would make sense since the app wasn't installed anymore. ``` [2023-09-11 18:02:24.555646] [STDOUT] stdout: [ +6 ms] Lost connection to device. Trying to connect again... [2023-09-11 18:02:24.556949] [STDOUT] stdout: [ +1 ms] executing: /opt/s/w/ir/x/w/recipe_cleanup/tmp53fs1szo/flutter sdk/bin/cache/artifacts/libimobiledevice/idevicesyslog -u 00008030-00144DA10185402E [2023-09-11 18:02:24.557323] [STDOUT] stdout: [ ] executing: script -t 0 /dev/null /opt/s/w/ir/x/w/recipe_cleanup/tmp53fs1szo/flutter sdk/bin/cache/artifacts/ios-deploy/ios-deploy --id 00008030-00144DA10185402E --bundle build/ios/iphoneos/Flutter Gallery.app --app_deltas build/ios/app-delta --uninstall --noinstall --debug --no-wifi --args --enable-dart-profiling --disable-vm-service-publication --enable-checked-mode --verify-entry-points [2023-09-11 18:02:24.578010] [STDOUT] stdout: [ +20 ms] [....] Waiting for iOS device to be connected [2023-09-11 18:02:24.712631] [STDOUT] stdout: [ +134 ms] [....] Using 00008030-00144DA10185402E (N104AP, iPhone 11, iphoneos, arm64e, 16.2, 20C65) a.k.a. 'iPhone 11'. [2023-09-11 18:02:24.712725] [STDOUT] stdout: [ ] ------ Uninstall phase ------ [2023-09-11 18:02:24.818293] [STDOUT] stdout: [ +105 ms] [ OK ] Uninstalled package with bundle id io.flutter.examples.gallery [2023-09-11 18:02:24.906833] [STDOUT] stdout: [ +88 ms] ------ Debug phase ------ [2023-09-11 18:02:24.906924] [STDOUT] stdout: [ ] Starting debug of 00008030-00144DA10185402E (N104AP, iPhone 11, iphoneos, arm64e, 16.2, 20C65) a.k.a. 'iPhone 11' connected through USB... [2023-09-11 18:02:25.285252] [STDOUT] stdout: [ +378 ms] [ 0%] Looking up developer disk image [2023-09-11 18:02:25.529937] [STDOUT] stdout: [ +244 ms] [ 90%] Mounting developer disk image [2023-09-11 18:02:25.545261] [STDOUT] stdout: [ +15 ms] [ 95%] Developer disk image already mounted [2023-09-11 18:02:25.587923] [STDOUT] stdout: [ +42 ms] Detected path to iOS debug symbols: "Symbol Path: /Users/swarming/Library/Developer/Xcode/iOS DeviceSupport/16.2 (20C65) arm64e/Symbols" [2023-09-11 18:02:25.857177] [STDOUT] stdout: [ +269 ms] Script started, output file is /dev/null [2023-09-11 18:02:25.857259] [STDOUT] stdout: [ ] Script done, output file is /dev/null [2023-09-11 18:02:25.857511] [STDOUT] stdout: [ ] ios-deploy exited with code 0 [2023-09-11 18:02:25.858066] [STDOUT] stderr: [ ] Could not run build/ios/iphoneos/Flutter Gallery.app on 00008030-00144DA10185402E. [2023-09-11 18:02:25.858130] [STDOUT] stderr: [ ] Try launching Xcode and selecting "Product > Run" to fix the problem: [2023-09-11 18:02:25.858214] [STDOUT] stderr: [ ] open ios/Runner.xcworkspace [2023-09-11 18:02:25.858537] [STDOUT] stdout: [ ] Installing and launching... (completed in 52.4s) [2023-09-11 18:02:25.858956] [STDOUT] stderr: [ ] Error launching application on iPhone 11. ```
Updated the fix (#134542), so need to wait again to see if it stops flaking |
[prod pool] flaky ratio for the past (up to) 100 commits between 2023-09-13 and 2023-09-19 is 2.22%. Flaky number: 2; total number: 90. Recent test runs: |
This has not flaked since workaround was added. I'm going to be cautiously optimistic and consider this fixed. For future, if this error reoccurs, see the following comment for a breakdown of the situation: |
This thread has been automatically locked since there has not been any recent activity after it was closed. If you are still experiencing a similar issue, please open a new bug, including the output of |
The post-submit test builder
Mac_ios hot_mode_dev_cycle_ios__benchmark
had a flaky ratio 2.00% for the past (up to) 100 commits, which is above our 2.00% threshold.One recent flaky example for a same commit: https://ci.chromium.org/ui/p/flutter/builders/prod/Mac_ios%20hot_mode_dev_cycle_ios__benchmark/4015
Commit: 00adf9a
Flaky builds:
https://ci.chromium.org/ui/p/flutter/builders/prod/Mac_ios%20hot_mode_dev_cycle_ios__benchmark/4015
https://ci.chromium.org/ui/p/flutter/builders/prod/Mac_ios%20hot_mode_dev_cycle_ios__benchmark/3995
Recent test runs:
https://flutter-dashboard.appspot.com/#/build?taskFilter=Mac_ios%20hot_mode_dev_cycle_ios__benchmark
Please follow https://github.com/flutter/flutter/wiki/Reducing-Test-Flakiness#fixing-flaky-tests to fix the flakiness and enable the test back after validating the fix (internal dashboard to validate: go/flutter_test_flakiness).
The text was updated successfully, but these errors were encountered: