Karl Stoney
Karl Stoney
That's valid, it's probably not worth a PR then given what you're suggesting is the processors are based on preference rather than accepted common stuff? I'll continue to maintain a...
So thinking about this a bit more @yurishkuro, maybe just adding the processors that are compatible with traces might be the compromise (rather than the whole world which wouldn't be...
+1, we need to be able to configure timeouts for any of the stages of an agentflow
Same issue here, can't sus this one out!
So this is caused by having `'--trace-events-enabled', '--trace-event-categories', 'v8'` where the doctor spawns the process. `v18.14.2` btw. Looks to be https://github.com/nodejs/node/issues/22042?
OK digging further this only happens in apps for us that include `@google-cloud/resource-manager`
Raised: https://github.com/googleapis/google-cloud-node/issues/4384 Not sure of the underlying cause however, hopefully the developers of that library will have a clue which helps others
Also raised https://github.com/nodejs/node/issues/48598, I think this is the root cause of the bug. In doctor, it causes a SIGABRT because of this line: https://github.com/clinicjs/node-clinic-doctor/blob/00749d83f04cb30ed24e91f29e89ad25e267ac33/injects/sampler.js#L22
interestingly this isn't a problem on node16, only 18
Sure, but it would be nice for users to be able to see the impact of custom authz requests in their traces