Summary

Test client backpressure overload retry when retryReads/Writes=false

Motivation

Review of the C driver discovered runCommand was retried on overload errors when retryReads=false or retryWrites=false, contradicting this spec requirement:

To retry runCommand, both retryWrites and retryReads MUST be enabled

Who is the affected end user?

Users upgrading to drivers with client backpressure.

How does this affect the end user?

Confused? But may result in unexpected retries.

How likely is it that this problem or use case will occur?

Likely? Identified reviewing C driver implementation of client backpressure retry (see comment)

If the problem does occur, what are the consequences and how severe are they?

Minor? Unexpected additional load on server?

Is this issue urgent?

I think this should be done as part of DRIVERS-3160 to ensure the spec requirement is tested.

Is this ticket required by a downstream team?

No?

Is this ticket only for tests?

Yes.

Acceptance Criteria

Test overload retry does not occur on runCommand when retryReads=false or retryWrites=false.