Intermittent connection abort from mock USS during INJECT request

[srikarBudda]

Describe the bug

During the recent test executions, the mock USS returns a connection error at the INJECT step. The request fails with Connection aborted.', ConnectionResetError(104, 'Connection reset by peer').
There is a similar ticket created but that ticket was for DELETE step which talks about including ConnectionResetError in the set of conditions that indicate a retryable query.

Screenshots

Additional context
Please find the report attached below.
You can find this issue at Scenario 268.

71b193fc-c936-4238-a3bb-cada9440f65b.zip

[BenjaminPelletier]

I have seen a number of instances like this where InterUSS products -- both mock_uss as in this case, and a DSS instance as in #1306 -- appear to reset the connection. As far as I know, this seems to be an issue with the hosting infrastructure. I do not believe we have seen any InterUSS product behavior that could be causing this issue, but there does seem to be some connection persistence issues (specifically, lost FIN packets) when using GCP infrastructure, perhaps related to load balancers. I would recommend raising this issue to the entity hosting mock_uss to see if they can resolve or trace the issue. If the issue does turn out to be related to some aspect of an InterUSS product's behavior, we are certainly interested to pursue further, but I think the first step for this particular issue is to pursue with hosting entity. If the hosting entity can provide additional information (such as internal logs for mock_uss when attempting to handler this request), we can investigate further.

[github-actions]

This issue has been automatically closed because there has been no response to our request for more information from the original author. With only the information that is currently in the issue, we don't have enough information to take action. Please reach out if you have or find the answers we need so that we can investigate further.

[davidmurphy1152]

For posterity: confirming closure appropriate as the available information is pointing more to a hosting issue rather than a software issue. We'll continue to monitor and if we bump on this again we'll raise in the appropriate venue for hosting/IT concerns.