Description
Expected Behaviour
The function logs should also contain an TraceID, which can be used to debug a specific function call.
Current Behaviour
Current logs shows timestamp, stderr/stdout, and function's name as prefix
Possible Solution
Modify 'bindLoggingPipe' signature and include an argument for (prefix/tag/traceId: string) in./executor/logging.go: bindLoggingPipe()
The tag can be parsed from HTTP Request, and have a fallback with default prefix from the environment variable.
Context
I am using OpenFaas in the backend and triggering async call to the function via S2S call from our API Server.
This can be considered as a user created Job, which might take few minutes / hours to complete.
User can come back any time, and check the status / logs to track the progress of the Job ( at the frontend ).
I am persisting the logs with ELK, but to cater this use case, I would have to create a custom logging logic for each function.
Since, the platform allows different user to create custom functions, I don't want the enduser to implement this required logic, and have a centralised provision for the same.
Your Environment
Kubernetes