Description

Summary of Changes

Increases timeoutMS in four CSOT ITERATION-mode cursor tests that were intermittently failing on sharded+auth+TLS CI variants.

What is the motivation for this change?

Every ITERATION-mode "refreshes the timeout" test works by proving a counter-factual: if the timer were not reset between operations, the cumulative block time would exceed timeoutMS and the cursor would error.

This requires satisfying two constraints simultaneously:

blockTimeMS < timeoutMS          # each individual op must succeed
N × blockTimeMS > timeoutMS      # total without reset must exceed the budget

The previous values left only 50 ms of margin per operation (timeoutMS - blockTimeMS = 50 ms). On fast topologies that is enough; on sharded+auth+TLS every round-trip carries extra overhead (TLS re-use, mongos routing) that regularly pushes an operation past the 50 ms window, causing a flaky timeouts.

History: #4278 previously doubled both blockTimeMS and timeoutMS (50 to 150 and 100 to 200) but kept the same 50 ms margin, so the underlying problem remained.

Double check the following

• Lint is passing (npm run check:lint)
• Self-review completed using the steps outlined here
• PR title follows the correct format: type(NODE-xxxx)[!]: description
  • Example: feat(NODE-1234)!: rewriting everything in coffeescript
• Changes are covered by tests
• New TODOs have a related JIRA ticket