> we will still have the inefficiency of submitting/updating one function 
> requiring the publishing of all assignments which can be large.

That's true.

> In the assignment topic each message will have a key that is the function id 
> and the value with be the assignments of its instances. And we can use topic 
> compaction to trim old entries.

yes, that could be a better approach. so, where this compaction will run? 
within function worker process?

[ Full content available at: 
https://github.com/apache/incubator-pulsar/pull/2474 ]
This message was relayed via gitbox.apache.org for [email protected]

Reply via email to