[Raptor] Test run failures aren't triggering a retry

RESOLVED WORKSFORME

Status

RESOLVED WORKSFORME
4 years ago
4 years ago

People

(Reporter: Eli, Assigned: Eli)

Tracking

({perf})

Firefox Tracking Flags

(Not tracked)

Details

(Assignee)

Description

4 years ago
In running Raptor tests on Jenkins, I noticed that a few times an application caused a test to timeout, but Raptor made no attempt to retry the test run. [1] contains a specific example.

[1] http://jenkins1.qa.scl3.mozilla.com/job/flame-kk-319.b2g-inbound.tinderbox.raptor.gaia/72/artifact/output.log
(Assignee)

Comment 1

4 years ago
Stack trace from raptor output in description:

Error: Test timeout exceeded 60000ms
    at Phase.timeoutError (/var/jenkins/1/workspace/flame-kk-319.b2g-inbound.tinderbox.raptor.gaia/node_modules/gaia-raptor/lib/suite/phase.js:50:22)
    at Phase.fail (/var/jenkins/1/workspace/flame-kk-319.b2g-inbound.tinderbox.raptor.gaia/node_modules/gaia-raptor/lib/suite/phase.js:233:12)
    at /var/jenkins/1/workspace/flame-kk-319.b2g-inbound.tinderbox.raptor.gaia/node_modules/gaia-raptor/lib/suite/phase.js:255:14

Phase.fail:phase.js 233:12 is the else portion of the following if:

  if (this.currentTry <= this.options.retries)

Since this.currentTry is set to 1 until after failure, and this.options.retries is defaulted to 1, not sure why this part isn't functioning correctly. Investigating.
(Assignee)

Comment 2

4 years ago
So far I have been unable to replicate this locally. I removed the fullyLoaded marker from Clock and retested by manually outputting a fullyLoaded log entry during priming, then letting the app hang during the test run. Raptor correctly did another retry. I nuked my node_modules as well but to no avail.

I'll leave this bug open for further investigation if something comes up.
(Assignee)

Updated

4 years ago
Status: ASSIGNED → RESOLVED
Last Resolved: 4 years ago
Resolution: --- → WORKSFORME
You need to log in before you can comment on or make changes to this bug.