OSX: intermittent failure: async-hooks/test-callback-error

[mhdawson]

• Version: master
• Platform: OSX
• Subsystem: async hooks

Failure in run for unrelated change:

https://ci.nodejs.org/job/node-test-commit-osx/12183/nodes=osx1010/console

not ok 76 async-hooks/test-callback-error
  ---
  duration_ms: 15.551
  severity: fail
  stack: |-
    start case 1
    end case 1: 114.761ms
    start case 2
    end case 2: 115.167ms
    start case 3
    end case 3: 8.221ms
    Error: test_callback_abort
        at ActivityCollector.initHooks.oninit.common.mustCall (/Users/iojs/build/workspace/node-test-commit-osx/nodes/osx1010/test/async-hooks/test-callback-error.js:36:45)
        at ActivityCollector.oninit (/Users/iojs/build/workspace/node-test-commit-osx/nodes/osx1010/test/common/index.js:509:15)
        at ActivityCollector._init (/Users/iojs/build/workspace/node-test-commit-osx/nodes/osx1010/test/async-hooks/init-hooks.js:182:10)
        at emitInitNative (async_hooks.js:446:43)
        at Object.emitInitScript [as emitInit] (async_hooks.js:349:3)
        at Object.<anonymous> (/Users/iojs/build/workspace/node-test-commit-osx/nodes/osx1010/test/async-hooks/test-callback-error.js:38:17)
        at Module._compile (module.js:549:30)
        at Object.Module._extensions..js (module.js:560:10)
        at Module.load (module.js:483:32)
        at tryModuleLoad (module.js:446:12)
     1: node::Abort() [/Users/iojs/build/workspace/node-test-commit-osx/nodes/osx1010/out/Release/node]
     2: node::Chdir(v8::FunctionCallbackInfo<v8::Value> const&) [/Users/iojs/build/workspace/node-test-commit-osx/nodes/osx1010/out/Release/node]
     3: v8::internal::FunctionCallbackArguments::Call(void (*)(v8::FunctionCallbackInfo<v8::Value> const&)) [/Users/iojs/build/workspace/node-test-commit-osx/nodes/osx1010/out/Release/node]
     4: v8::internal::MaybeHandle<v8::internal::Object> v8::internal::(anonymous namespace)::HandleApiCallHelper<false>(v8::internal::Isolate*, v8::internal::Handle<v8::internal::HeapObject>, v8::internal::Handle<v8::internal::HeapObject>, v8::internal::Handle<v8::internal::FunctionTemplateInfo>, v8::internal::Handle<v8::internal::Object>, v8::internal::BuiltinArguments) [/Users/iojs/build/workspace/node-test-commit-osx/nodes/osx1010/out/Release/node]
     5: v8::internal::Builtin_Impl_HandleApiCall(v8::internal::BuiltinArguments, v8::internal::Isolate*) [/Users/iojs/build/workspace/node-test-commit-osx/nodes/osx1010/out/Release/node]
     6: 0x31be80046fd

[mhdawson]

stress run https://ci.nodejs.org/job/node-stress-single-test/1404/

[Trott]

Looks like the macOS machine got re-provisioned but the step about disabling ReportCrash probably didn't happen. See #13527 (comment).

@nodejs/build Can someone run those commands? (It needs to be from GUI login unfortunately, so you can't just SSH in.) And also if it's possible to automate those two commands somehow in Ansible or something, that would be great, but I'm not sure how difficult that is given the GUI login requirement.

[Trott]

Hmmm...actually, this is different it seems. This is a stack crash whereas the previous issue was a timeout, so maybe ReportCrash isn't related here after all...

[Trott]

It seems to be crashing after successfully running all three test cases.
¯\(ツ)/¯ @nodejs/async_hooks

[bcoe]

@Trott @mhdawson I was seeing this same behavior on my local OSX tests, and running:

launchctl unload -w /System/Library/LaunchAgents/com.apple.ReportCrash.plist
sudo launchctl unload -w /System/Library/LaunchDaemons/com.apple.ReportCrash.Root.plist

did seem to do the trick; Could you just have Jenkins run that as a pre-test step on OSX (kind of dumb, but works).

[Trott]

@bcoe Thanks for the information! The commands were run by @rvagg and that seems to have fixed it on CI. So yes, that did the trick after all. Hooray!

Rather than running it in Jenkins, I think ansible-minded people on @nodejs/build can hopefully add it to the macOS ansible playbook. Or maybe they'll decide Jenkins is the better place. That's a discussion for https://github.com/nodejs/build. In the meantime, problem solved (for now, at least) so I'm going to close this!

[mhdawson]

If ifs a one time thing it should go into ansible. @gdams we need to make sure we include this in the ansible scripts for OSX.