roachtest: fix costfuzz log on error while running perturbed statement #81102
Conversation
Also log the unperturbed statement when the costfuzz test encounters an error executing the perturbed statement. This can be fixed by hand, but it's easier if the costfuzz log is correct. Informs: cockroachdb#81032 Release note: None
|
This change is |
There was a problem hiding this comment.
Reviewed 1 of 1 files at r1, all commit messages.
Reviewable status:complete! 1 of 0 LGTMs obtained (waiting on @msirek)
|
TFTR! bors r=rytaft |
There was a problem hiding this comment.
Reviewable status:
pkg/cmd/roachtest/tests/costfuzz.go line 226 at r1 (raw file):
rows2, err := conn.Query(stmt) if err != nil { logStmt(stmt)
Does calling this function twice cause different behavior on the 2nd call? It's also called below.
|
Build failed (retrying...): |
There was a problem hiding this comment.
Reviewable status:
pkg/cmd/roachtest/tests/costfuzz.go line 226 at r1 (raw file):
Previously, msirek (Mark Sirek) wrote…
Does calling this function twice cause different behavior on the 2nd call? It's also called below.
Calling this function twice simply logs the statement twice. (The log is supposed to be a record of statements executed during this run. We executed this statement twice, once normally and once after setting the perturbation seed, so we need the log to reflect that for reproducibility.)
There was a problem hiding this comment.
Reviewable status:
pkg/cmd/roachtest/tests/costfuzz.go line 226 at r1 (raw file):
Previously, michae2 (Michael Erickson) wrote…
Calling this function twice simply logs the statement twice. (The log is supposed to be a record of statements executed during this run. We executed this statement twice, once normally and once after setting the perturbation seed, so we need the log to reflect that for reproducibility.)
Right, of course. Thanks!
|
Build succeeded: |
Also log the unperturbed statement when the costfuzz test encounters an
error executing the perturbed statement.
This can be fixed by hand, but it's easier if the costfuzz log is
correct.
Informs: #81032
Release note: None