Make sure that snapshot reads with atClusterTime in testDoNotRejectReadsAfterMigrationAbortedAfterReachingBlockTimestamp and testDoNotRejectReadsAfterMigrationAbortedAfterReachingBlockTimestamp get assigned read timestamp >= blockTimestamp.
Make testDoNotRejectReadsAfterMigrationAbortedAfterReachingBlockTimestamp test that if the migration aborts after returnAfterReachingTimestamp (blockTimestamp) is reached, the recipient keeps rejecting reads with atClusterTime < returnAfterReachingTimestamp (blockTimestamp) until the state doc is garbage collected. Currently, it tests that the recipient stops rejecting reads after the migration aborts.
Bug
Major - P3