Currently when sending telemetry to the server, telemetry client will retry telemetry flushing attempt up to 3 times with no timeout for the outgoing request, which might lead to very big timeouts in some certain actions happening in mongosh, for example trying to exit with a flush attempts failing with a timeout (8 minutes in one reported case).

{“t”:{“$date”:“2022-06-14T01:31:15.265Z”},“s”:“I”,“c”:“MONGOSH”,“id”:1000000011,“ctx”:“shell-api”,“msg”:“Performed API call”,“attr”:{“method”:“countDocuments”,“class”:“Collection”,“db”:“jhotel”,“coll”:“hotelInfo”,“arguments”:{“options”:{}}}}
{“t”:{“$date”:“2022-06-14T01:39:46.049Z”},“s”:“I”,“c”:“MONGOSH”,“id”:1000000045,“ctx”:“analytics”,“msg”:“Flushed outstanding data”,“attr”:{“flushError”:“connect ETIMEDOUT 44.242.123.48:443",“flushDuration”:510755}}

We want to provide a default configuration for the telemetry that will introduce a request timeout of 1 second and disable the retries completely (see thread for some additional context)