Bug
- [SERVER-3528] - memory footprint of JS context keeps increasing as it is used
- [SERVER-3661] - Specifying includeLocs returns objects instead of arrays when arrays are used for locations
- [SERVER-3715] - re-adding member to replica set without keyFile results in secondary node
- [SERVER-3837] - BuildInfo shouldn't require auth
- [SERVER-3842] - Multiple mongos can interfere in initial sharding
- [SERVER-3899] - Initial sync should check occaisonally if it can stop and become secondary
- [SERVER-4020] - bad serverID set in setShardVersion
- [SERVER-4114] - If a shard contacts another shard for the first time for a MapReduce, the sharding connection hook won't be set
- [SERVER-4118] - mongos causes dos by opening a ton of connections
- [SERVER-4137] - Need to catch DBException in message_server_port
- [SERVER-4160] - Make sure members that build indexes sync from members with indexes
- [SERVER-4167] - Sharding info not communicated to all mongos servers - results in Assertion Errors
- [SERVER-4171] - WritebackListener should force detection of newly sharded collection
- [SERVER-4178] - Shared client doesn't compile - references code for server side components.
- [SERVER-4196] - Count command on mongos can get into infinite loop
- [SERVER-4228] - Segfault under load
- [SERVER-4240] - make 2.0.1 mongos nightly compatible with 2.0.1 mongod
- [SERVER-4250] - db.currentOp and db.killOp on mongos work without needing to auth()
- [SERVER-4270] - master/slave initial sync bug backport
- [SERVER-4290] - Don't dispose of connections after GLE
- [SERVER-4296] - don't allow too many splitVectors in parallel from splitIfShould
- [SERVER-4324] - new checking of initial shard version catches stale replica set info
- [SERVER-4331] - conneciton/memory leak if onCreate or onHandedOut hooks fail
- [SERVER-4350] - Segmentation fault on replica recovery
- [SERVER-4405] - replSetStepDown must close socket on step down.
- [SERVER-4430] - Initial sync reclone asserts on correctly applied oplog entry
- [SERVER-4465] - counts with large numbers of shards can error with too many retries
Task
- [SERVER-4209] - Remove warning and stack trace when adding shard sub-connection
Improvement
- [SERVER-3829] - scons flag to use each library externally --use-system-(prcre|boost) etc..
- [SERVER-4063] - make leak warning take number of connections info consideration WAS: I would like a way to set the threshold for the memory leak warning message
- [SERVER-4242] - geShardDistribution in 2.0.2
- [SERVER-4396] - hit config server only once when loading a chunk meta data certain version
Edit/Copy Release Notes
The text area below allows the project release notes to be edited and copied to another document.