You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
We need to track who, what, and how work gets done on the protocol. This means we need real time indexed insight into the amount of work requests happening on the network, who is completing them, and what errors are occurring. We have implemented a lightweight package in this PR: #513
We now need to implement this package in our codebase.
Acceptance criteria
The events we should track are as follows:
Work Distribution Event:
Track when work is distributed to a worker
Include information about whether it's a remote or local worker
Capture the work type and peer ID
Work Completion Event:
Record when a work item is successfully completed
Include the work type and peer ID
Add a flag indicating whether the work was successful
Worker Failure Event:
Log when a worker fails to complete a task
Include the work type, peer ID, and error message
Work Execution Start Event:
Track when work execution begins
Include the work type and whether it's a local or remote execution
Work Execution Timeout Event:
Record when work execution times out
Include the work type and the timeout duration
Remote Worker Connection Event:
Log when a connection is established with a remote worker
Include the remote worker's peer ID
Stream Creation Event:
Track when a new stream is created for communication with a remote worker
Include the remote worker's peer ID and the protocol used
Work Request Serialization Event:
Record when a work request is serialized for transmission
Include the work type and the size of the serialized data
Work Response Deserialization Event:
Log when a work response is deserialized after reception
Include the work type and whether deserialization was successful
Local Worker Fallback Event:
Track when the system falls back to using a local worker
Include the reason for fallback (e.g., no remote workers available, all remote workers failed)
The text was updated successfully, but these errors were encountered:
There was an issue with sending tracking for the local node because we were not populating the AddrsInfo for it. Adding that to the local solves the runtime error.
Problem
We need to track who, what, and how work gets done on the protocol. This means we need real time indexed insight into the amount of work requests happening on the network, who is completing them, and what errors are occurring. We have implemented a lightweight package in this PR: #513
We now need to implement this package in our codebase.
Acceptance criteria
The events we should track are as follows:
Work Distribution Event:
Work Completion Event:
Worker Failure Event:
Work Execution Start Event:
Work Execution Timeout Event:
Remote Worker Connection Event:
Stream Creation Event:
Work Request Serialization Event:
Work Response Deserialization Event:
Local Worker Fallback Event:
The text was updated successfully, but these errors were encountered: