Currently, we have no metrics on the task queues, therefore we don't yet know how long our proofs take to compute, how much of them there are, etc.
Bullmq offers a nice integration: https://docs.bullmq.io/guide/telemetry
However, there are different combinations of things we have to cover:
- Consolelogger + Localtaskqueue
- Consolelogger + Bullmq
- OpenTelemetry + Localtaskqueue
- OpenTelemetry + Bullmq
Using the tracer inside Localtaskqueue is rather straight-forward.
But then, we need to build a interface for the Bullmq Tracer (https://github.com/taskforcesh/bullmq/blob/master/src/interfaces/telemetry.ts)
that takes care of mapping our tracer interface to the bullmq tracer interface, so that we can dynamically inject the tracer into the configured bullmq module
Currently, we have no metrics on the task queues, therefore we don't yet know how long our proofs take to compute, how much of them there are, etc.
Bullmq offers a nice integration: https://docs.bullmq.io/guide/telemetry
However, there are different combinations of things we have to cover:
Using the tracer inside Localtaskqueue is rather straight-forward.
But then, we need to build a interface for the Bullmq Tracer (https://github.com/taskforcesh/bullmq/blob/master/src/interfaces/telemetry.ts)
that takes care of mapping our tracer interface to the bullmq tracer interface, so that we can dynamically inject the tracer into the configured bullmq module