Correct way to disconnect from persistent subscription

I modified our code and added stop as you had recommended but that is crashing our service. I was also able to replicate the issue with the test program I sent you.

Hi Anand,

I do apologise for not mentioning this in my last comment, but calling Stop() will not help fix this issue.

We will be working on a fix for this after the release of 4.0.2.

For now, there are a few options for working around this:

  1. The simplest is to find the source of the slow messages on your servers

i. What are the specs of your VM?

ii. What kind of disks are you running on?

iii. What kind of load are you putting on your Event Store when you see these issues?

  1. You could reduce the number of occurrences of this issue by increasing the operation timeout on your connection settings. However, from the logs that you provided, there were some cases where the queues were taking more than 300 seconds to process a message. So in order for this to work, this timeout value would need to be set to a very high number.

Are there any cons in increasing the operation timeout?

Our production environment should have better specs than our internal test servers. So, 1(i) and 1(ii) should not be an issue. For 1(iii), the issue occurs when no one is using eventstore over the weekend. Never seen it occur when we send and receive events without huge periods of inactivity.

300 seconds is not reasonable in any environment.