Our solution consists of a series of Azure functions that have EventHub triggers that subscribe to EventHubs. De functions form a chain and messages get passed from function to function via eventHubs.
We used to get Socket Exceptions when sending these messages, which we’ve fixed by implementing our own retry code.
Since about March / April this year we’ve been getting different kind of socket exceptions, which appear to be coming from the Eventhub processor (so the receiving side, NOT from sending side)
Here’s a snippet of the Stacktrace:
[{"assembly":"Microsoft.Azure.EventHubs.Processor, Version=4.3.2.0, Culture=neutral, PublicKeyToken=7e34167dcc6d6d8c","method":"Microsoft.Azure.EventHubs.Processor.AzureStorageCheckpointLeaseManager+<ReleaseLeaseCoreAsync>d__37.MoveNext","level":0,"line":0}]
We are not losing any messages as far as we can tell, and the code seems to be self-correcting. The error does appear a few times a day.
We’ve contacted Microsoft, but they can’t figure out what’s going on, unless we set one of our Solutions on a public plan, so that they can do more with that. Unfortunately, our company, for security reasons, does not allow this.
Have you seen the problem before and have you found any solution to this problem?