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 are currently considering using this plugin to integrate tracing from requests to Kong with the rest of our system. The readme has the following warning it in now.
This is proof of concept code, not production ready
What tasks/features are lacking currently for this to be production ready? If we could make that clear and track them here, people using this plugin and others in the community can more easily help make it production ready.
One thing I think I see by reading the code (though I'm not familiar with Kong or Lua at all) is it looks like it is sending the span data on each request instead of batching. I don't know how easy or possible it is, but batching sending spans could improve performance of reporting spans to Zipkin.
The text was updated successfully, but these errors were encountered:
The warning is just there because I've personally only used Zipkin in proof-of-concept / prototype scenarios so far. I'm probably not going to get to a production deployment for several months.
I'm not aware of any issues preventing production use. Batching would be nice, but not sure how feasible it is (might require writing to the Postgres/Cassandra backend?). The only other nice-to-have I can think of is reporting the response code to Zipkin
We are currently considering using this plugin to integrate tracing from requests to Kong with the rest of our system. The readme has the following warning it in now.
What tasks/features are lacking currently for this to be production ready? If we could make that clear and track them here, people using this plugin and others in the community can more easily help make it production ready.
One thing I think I see by reading the code (though I'm not familiar with Kong or Lua at all) is it looks like it is sending the span data on each request instead of batching. I don't know how easy or possible it is, but batching sending spans could improve performance of reporting spans to Zipkin.
The text was updated successfully, but these errors were encountered: