Refine the retry logic #93
Description
What feature do you want to see added?
Currently we are using a very simple 5s polling. While this is simple, this also may cause DDOS if a lot of plugins start polling the server during an outage.
It'd be nice to make this a bit smarter and recognize "normal operations" from "errors" so that errors could use exponential backoff.
See also #86
Upstream changes
No response
Are you interested in contributing this feature?
No response