Skip to content

Longer timeout value instead of multiple retries with shorter timeout

Mateusz Żebrak requested to merge mzebrak/no-retry-timeout into develop

Because it looks there is no sense to retry on timeout, just have a longer timeout. That's because if something is under heavy load, and we have a timeout set for e.g. 6s while it potentially could answer in 7s, we won't be able to correctly communicate and still CommunicationTimeoutError will be send after the same time as longer timeout set (6s * 5 (retries) = 30s)

So with this change we should have less timeout errors and the time for TimeoutError notification to display will remain the same.

Merge request reports