| Builder |
Machine Type |
P25 Duration In Minutes |
Average Duration In Minutes |
P75 Duration In Minutes |
Total Builds |
Duration Bucket |
| Mac tool_integration_tests_4_5 |
n1-highcpu-8 |
33.62 |
34.72 |
34.97 |
88 |
30-35 |
| Mac tool_integration_tests_1_5 |
n1-highcpu-8 |
30.68 |
31.45 |
31.73 |
90 |
30-35 |
| Mac tool_integration_tests_5_5 |
n1-highcpu-8 |
29.48 |
30.40 |
30.87 |
89 |
30-35 |
| Mac tool_integration_tests_3_5 |
n1-highcpu-8 |
20.30 |
20.68 |
21.53 |
95 |
20-25 |
| Mac tool_integration_tests_2_5 |
n1-highcpu-8 |
16.03 |
16.76 |
16.90 |
83 |
15-20 |
All 5 Mac tool_integration_tests_* are regularly taking more than the targeted 10-15 minutes. While we could be smarter about bucketing, we could also take the easier road and just increase the number of subshards. Mac currently has 5 shards. For comparison, Windows has 10 and Linux has 7. Increase the number of shards until each are under 15 minutes.
If there are still outliers then consider smarter organization, like pulling out the longest running ones into separate shards.
Currently, the try pool capacity for this test is 129 bots and the prod pool has 56 bots.
Set like this:
|
- name: Mac tool_integration_tests_1_5 |
All 5
Mac tool_integration_tests_*are regularly taking more than the targeted 10-15 minutes. While we could be smarter about bucketing, we could also take the easier road and just increase the number of subshards. Mac currently has 5 shards. For comparison, Windows has 10 and Linux has 7. Increase the number of shards until each are under 15 minutes.If there are still outliers then consider smarter organization, like pulling out the longest running ones into separate shards.
Currently, the try pool capacity for this test is 129 bots and the prod pool has 56 bots.
Set like this:
flutter/.ci.yaml
Line 4576 in bcabd8e
flutter/.ci.yaml
Line 4591 in bcabd8e