PYTHON-4730 Failing Async Bulk Tests

[sleepyStick]

Honestly not sure what changed since this PR that would cause the tests to fail, because they weren't failing when I merged them...?

[ShaneHarvey]

Looks like evergreen never actually ran on the final commit in #1827 so that would explain why the tests "passed".

[sleepyStick]

ahhh that makes sense, sorry for not noticing that earlier!

[ShaneHarvey]

 [2024/09/04 16:20:57.567] FAILURE: pymongo.errors.ServerSelectionTimeoutError: localhost:27017: connection closed (configured timeouts: socketTimeoutMS: 20000.0ms, connectTimeoutMS: 20000.0ms), Timeout: 30s, Topology Description: <TopologyDescription id: 66d8ea2454b87254ccbe234d, topology_type: Single, servers: [<ServerDescription ('localhost', 27017) server_type: Unknown, rtt: None, error=AutoReconnect('localhost:27017: connection closed (configured timeouts: socketTimeoutMS: 20000.0ms, connectTimeoutMS: 20000.0ms)')>]> ()
 [2024/09/04 16:20:57.567] self = <test.asynchronous.test_bulk.AsyncTestBulkWriteConcern testMethod=test_write_concern_failure_ordered>
 [2024/09/04 16:20:57.567]     @async_client_context.require_replica_set
 [2024/09/04 16:20:57.567]     @async_client_context.require_secondaries_count(1)
 [2024/09/04 16:20:57.567]     async def test_write_concern_failure_ordered(self):
 [2024/09/04 16:20:57.567]         # Ensure we don't raise on wnote.
 [2024/09/04 16:20:57.567]         coll_ww = self.coll.with_options(write_concern=WriteConcern(w=self.w))
 [2024/09/04 16:20:57.567]         result = await coll_ww.bulk_write([DeleteOne({"something": "that does no exist"})])
 [2024/09/04 16:20:57.567]         self.assertTrue(result.acknowledged)
 [2024/09/04 16:20:57.567]     
 [2024/09/04 16:20:57.567]         requests: list[Any] = [InsertOne({"a": 1}), InsertOne({"a": 2})]
 [2024/09/04 16:20:57.567]         # Replication wtimeout is a 'soft' error.
 [2024/09/04 16:20:57.567]         # It shouldn't stop batch processing.
 [2024/09/04 16:20:57.567]         try:
 [2024/09/04 16:20:57.567] >           await self.cause_wtimeout(requests, ordered=True)
 [2024/09/04 16:20:57.567] test/asynchronous/test_bulk.py:991: 
 [2024/09/04 16:20:57.567] _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ 
 [2024/09/04 16:20:57.567] test/asynchronous/test_bulk.py:971: in cause_wtimeout
 [2024/09/04 16:20:57.567]     await self.secondary.admin.command("configureFailPoint", "rsSyncApplyStop", mode="alwaysOn")
 [2024/09/04 16:20:57.567] pymongo/_csot.py:120: in csot_wrapper
 [2024/09/04 16:20:57.567]     return func(self, *args, **kwargs)
 [2024/09/04 16:20:57.567] pymongo/synchronous/database.py:904: in command
 [2024/09/04 16:20:57.567]     with self._client._conn_for_reads(read_preference, session, operation=command_name) as (
 [2024/09/04 16:20:57.567] pymongo/synchronous/mongo_client.py:1675: in _conn_for_reads
 [2024/09/04 16:20:57.567]     server = self._select_server(read_preference, session, operation)
 [2024/09/04 16:20:57.567] pymongo/synchronous/mongo_client.py:1623: in _select_server
 [2024/09/04 16:20:57.567]     server = topology.select_server(
 [2024/09/04 16:20:57.567] pymongo/synchronous/topology.py:396: in select_server
 [2024/09/04 16:20:57.567]     server = self._select_server(
 [2024/09/04 16:20:57.567] pymongo/synchronous/topology.py:374: in _select_server
 [2024/09/04 16:20:57.567]     servers = self.select_servers(
 [2024/09/04 16:20:57.567] pymongo/synchronous/topology.py:281: in select_servers
 [2024/09/04 16:20:57.567]     server_descriptions = self._select_servers_loop(
 [2024/09/04 16:20:57.567] _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ 
 [2024/09/04 16:20:57.567] self = <Topology <TopologyDescription id: 66d8ea2454b87254ccbe234d, topology_type: Single, servers: [<ServerDescription ('loc...localhost:27017: connection closed (configured timeouts: socketTimeoutMS: 20000.0ms, connectTimeoutMS: 20000.0ms)')>]>>

self.secondary is mistakenly created as a sync MongoClient instead of AsyncMongoClient.

[ShaneHarvey]

Totally not your fault. Evergreen must have failed to schedule the PR run somehow. That almost never happens so it's not surprising that we missed it.

[sleepyStick]

good eye! I've been trying to figure out what went wrong for the better part of the last hour or so! >.< thanks!

[blink1073]

Yeah EVG had an outage today