Add retryReads URI option that defaults to True.
Supported read operations will be retried once after transient
network, election, and shutdown errors on MongoDB 3.6+.
Supported operations are:
listCollections, listIndexes, and listDatabases
distinct
count, estimated_document_count, count_documents
aggregate (not including $out)
find (only for the initial find command, getMore commands are not
retried).
ChangeStreams: watch (initial aggregate command).
GridFS read APIs.
Test changes:
Add retryable reads spec test runner.
Disable retryable reads in network error tests.
Move exhaust getMore out of Cursor.
Move cursor command response decoding into Server so that all command
listener events can be published in the same location.
Transient errors inside transaction unpins the session.
Add brief docs about sharded transactions and add 3.9 changelog.
Tests changes:
Add spec tests for sharded transaction recoveryToken.
Speed up txn tests by reducing SDAM waiting time after a network error.
Remove outdated test workaround for killAllSessions.
Use _OpReply class instead of passing bytes around.
Remove unnecessary operation argument to receive message.
Move _first_batch to message.py to avoid circular import.
The method had checked whether any SON manipulators were configured, and
raised an error if so. But users should understand that SON manipulators
can't affect raw batches, and checking for them raises a deprecation
warning.
Rename find_raw to find_raw_batches, and add aggregate_raw_batches.
Rename RawBSONCursor and RawBSONCommandCursor to RawBatchCursor and
RawBatchCommandCursor.
This commit implements the following simplifications and improvements:
- The Subscriber ABC has been renamed to CommandListener
- The subscribe function has been renamed to register
- The COMMAND constant has been removed
- The get_subscribers function has been removed
- Command listeners can now be registered per MongoClient instance
instead of just globally
We weren't including the time to decode responses in
a number of places. This commit also fixes an issue with
handling "cursor not found" errors from mongos versions
older than 2.4.
This commit adds support for monitoring queries and getMore
operations, including getMore operations for command cursors
(e.g. aggregate cursor getMore). It also fixes a bug in duration
calculation.
Even with this change, "next" can raise StopIteration even though
"alive" is True. For example if batch size is 2 and there are 4
documents in the result set, then after the 4th document "alive" is True
but "next" raises StopIteration.
MongoClient now supports all MongoReplicaSetClient's features.
Fix bugs and add features in the new MongoClient to bring it up
to spec. MongoReplicaSetClient is now a deprecated stub that
subclasses MongoClient.
Introduce new errors, NetworkTimeout and NotMasterError, to
communicate events that MongoClient must handle specially.
Replace MongoClient with an implementation that relies on Cluster and Server. The new MongoClient takes over MongoReplicaSetClient's responsibilities.
Authentication, secondary-pinning, and Mongos high-availability are broken and will be reimplemented in a future commit. RS tests are temporarily disabled.
