Reputation: 147
As per my business case, the processor application which processing the EventHub encounters any transient faults(429,449,503 etc) it should be retried as many times as it takes to succeed with exponential back-off in order to avoid data loss.
Does the EventHub lease will expire as the processor thread may go for longer sleep during retry's waiting period in exponential back-off approach?
Is above approach is recommended? if not how the data loss can be avoided during 429s or other transients faults?
Note: throwing message back to the Eventhub after few retries, is not an option in my case as it hampers the message ordering.
Upvotes: 1
Views: 521