x/build: linux-amd64-race TryBot straggling #32032
Open
Milestone
Comments
All the timing data is in BigQuery. That almost always reveals the mystery. |
Looks like the @rsc's recent suffixarray changes made it slower? https://console.cloud.google.com/bigquery?sq=872405196845:3b455f3831c54226ba07eda8cea81247 ... SELECT
Builder, Event, APPROX_QUANTILES(Seconds, 100)[OFFSET(50)] as MedianSec, COUNT(*) as N
FROM
builds.Spans
WHERE
Error = ''
AND Builder = 'linux-amd64-race'
AND StartTime > TIMESTAMP_SUB(CURRENT_TIMESTAMP(), INTERVAL 24*7 HOUR)
AND Repo = "go"
And IsTry = true
GROUP BY 1, 2
ORDER BY 3 DESC
|
But suffixarray seems fast on linux with -race locally, even before Russ's recent commit. |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
In the TryBot run for https://golang.org/cl/176580 at Patchset 4, most of the TryBots ran relatively quickly (less than 10m), but the
linux-amd64-race
builder took much longer (over 20m).I'm not sure exactly why it was so slow, but it leads me to believe that there may be something wrong with the sharding for that builder.
(CC @bradfitz @dmitshur @andybons)
The text was updated successfully, but these errors were encountered: