wikimedia
/
parsoid
/
885
/
2
85%
master: 85%
master: 85%
|
Ran
|
Files
80
|
Run time
4s
|
Badge
Embed ▾
README BADGES
|
coverage: 82.6%. Remained the same
push
travis-ci
subbussFollow up to 4d06e1ba * @Inez noticed that ctrl-c sends SIGINT to all processes in the group, causing workers to race to disconnect. This could put a worker in the "disconnected" state, making `worker.disconnect()` throw and preventing the timeout from being cleared. When it fires, we get a "kill ESRCH". * Cluster in node v0.8 doesn't have a `worker.kill()` method, so we use the internal worker.process.kill instead. This is what node does anyways and is safer than the `process.kill()` in the timeout, so we use it for both. * Finally, we prevent killed workers from starting back up on exit. That's something that was missed in the service-runner port. Change-Id: I61a8f034c
12879 of 15592 relevant lines covered (82.6%)
20379.79 hits per line
