Features of the latest MongoDB version

MongoDB 2.4 has been released. This version adds some new features, such as text search, hash-based sharding, better geospatial functions, support for geojson, and performance and tools. We also discussed the next road map of MongoDB with 10gen.

Some key improvements are as follows:

• Text Search is introduced. Now it is a test function that supports stem and break words in 15 languages.
• Hash-based sharding, which is used for cross-propagation of data, cannot easily predict any natural partition keys.
• Geospatial indexing and geojson support
• Security improvement-new modular identity authentication system, integration with Kerberos, and Role-Based Access Control
• Some performance improvements, the most important of which is the improvement in specialized scenarios such as counting and aggregation.
• V8 is used as the default JavaScript Engine in Mongo shell (instead of spidermonkey), which improves the performance and concurrency of JavaScript-based operations.
• Other metrics used to monitor the cluster status

10gen also introduced a MongoDB Enterprise Edition released with version 2.4.

To learn more about new features and future plans, we contacted Kelly stirman, product marketing director of 10gen.

Kelly explains why collection-level locks may be meaningless to MongoDB:

Version 2.2 has improved the lock and brought substantial benefits to the write throughput by reducing the lock competition. David mytton wrote a very good article on this topic.

MongoDB 2.0 does not contain any extra-granularity locks except the improvements provided in versions 2.2 and 2.4. We are considering introducing document-level locks in version 2.6. The lock generation upgrade is sufficient, and the set-Level Lock may not lead to any other important improvement. Therefore, the document-Level Lock may be in the next direction.

Kelly said:

When will range-based sharding be used? If your applicationProgramWhen you request data based on a shard key range, these queries are routed to the appropriate shard. Generally, there is only one shard. In special cases, there may be some shards. In a system that uses hash sharding, the same query routes requests to more shards, possibly all shards. Ideally, we want the query to be routed to a single shard or as few shards as possible, because this scalability is better than routing all the queries to all shards. Therefore, if you fully understand your data and queries, range-based sharding may be the best choice.

In MongoDB 2.4, the counting speed can be increased by up to 20 times, and the aggregation framework is faster than 3 to 5 times on average. Kelly explained that the increase in counting performance has benefited from some improvements in MongoDB's B-tree traversal performance-index-based low base count is the biggest improvement. The performance improvement of the aggregation framework is actually caused by many small changes made to the internal implementation of MongoDB, which bring great benefits together.

Which of the following enterprise features will be added:

MongoDB 2.4 has implemented some major initiatives in the security and monitoring fields, but we still have many plans for future versions. We will perform security control in three dimensions: authentication, authorization, and audit. In the future, MongoDB will continue to focus on these fields, and we will continue to enhance the tools we provide for MongoDB. MongoDB Monitoring Service (MMS) in MongoDBCommunityIt is already very popular with over 15,000 users and continues to grow. We will continue to invest in MMs to provide free cloud tools and local products as part of our enterprise subscription service.

For more information about the new features of MongoDB 2.4, read the release notes and overview.

View Original English text:MongoDB gets better security, text search, performance improvements-What's next?