Increase sharding for interactive_ui_tests on linux tsan CQ builder
This is testing for a hypothesis about how we can approach efficiently
sharding our test suites. This suite isn't significantly slowing
builds down; it's in the top 10 slowest tests (roughly), but isn't the
slowest test as far as I can tell.
This should result in roughly 30 more machines being used by the
builder, which should be very easy handled by robocrop's auto-scaling.
Bug: 1179567
Change-Id: I25a22ccdbd6f597f5504b4d5e9db151b87d5be16
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/2703713
Reviewed-by: Erik Staab <[email protected]>
Reviewed-by: Gary Tong <[email protected]>
Commit-Queue: Stephen Martinis <[email protected]>
Cr-Commit-Position: refs/heads/master@{#856950}
diff --git a/testing/buildbot/test_suite_exceptions.pyl b/testing/buildbot/test_suite_exceptions.pyl
index 03ae811..9c37eea 100644
--- a/testing/buildbot/test_suite_exceptions.pyl
+++ b/testing/buildbot/test_suite_exceptions.pyl
@@ -1513,7 +1513,7 @@
'Linux TSan Tests': {
# These are slow on the TSan bots for some reason, crbug.com/794372
'swarming': {
- 'shards': 20,
+ 'shards': 32, # Adjusted for testing, see https://crbug.com/1179567
},
},
'ToTLinuxTSan': {