rare test failure in CI: "tracing integration works" #46
Description
We sometimes (but rarely) get this test hang/error when running in CI.
I haven't yet seen it in local dev.
> @elastic/ecs-winston-format@0.3.0 test /var/lib/jenkins/workspace/nodejs_ecs-logging-js-mbp_master/src/go.elastic.co/apm/ecs-logging-js/loggers/winston
> standard && tap --100 --timeout ${TAP_TIMEOUT:-10} test/*.test.js
TAP version 13
not ok 1 - test/apm.test.js # time=300096.871ms
---
env: {}
file: test/apm.test.js
timeout: 300000
command: /usr/local/bin/node
args:
- -r
- /var/lib/jenkins/workspace/nodejs_ecs-logging-js-mbp_master/src/go.elastic.co/apm/ecs-logging-js/loggers/winston/node_modules/esm/esm.js
- test/apm.test.js
stdio:
- 0
- pipe
- 2
cwd: /var/lib/jenkins/workspace/nodejs_ecs-logging-js-mbp_master/src/go.elastic.co/apm/ecs-logging-js/loggers/winston
failures:
- tapError: no plan
exitCode: null
signal: SIGTERM
...
{
# Subtest: tracing integration works
ok 1 - should not error
ok 2 - apmServerUrl: http://localhost:34201
ok 3 - should be equal
ok 4 - first listening log line has "address"
ok 5 - should not error
ok 6 - appUrl: http://localhost:38745
ok 7 - should not error
ok 8 - logObj is ECS valid
ok 9 - should be equal
ok 10 - should be equal
ok 11 - should be equal
ok 12 - empty stderr from app
ok 13 - app exited 0
ok 14 - traceObjs[0] is metadata
ok 15 - traceObjs[1] is transaction
ok 16 - traceObjs[2] is span
ok 17 - trace.id matches
ok 18 - transaction.id matches
ok 19 - span.id matches
ok 20 - should be equal
ok 21 - should be equal
not ok 22 - timeout!
---
signal: SIGTERM
handles:
- type: Server
events:
- request
- connection
connectionKey: 6::::0
expired: TAP
stack: |
emit (node_modules/signal-exit/index.js:78:11)
process.listener (node_modules/signal-exit/index.js:92:7)
test: tracing integration works
...
1..22
# failed 1 of 22 tests
not ok 1 - tracing integration works # time=297190.283ms
# test count(1) != plan(null)
# failed 2 test
}