Senior Solutions Architect, MongoDB Inc. What’s New in MongoDB 2.6 Massimo Brignoli Senior Solutions Architect, MongoDB Inc.

Recent Release History 2.2 Aug ‘12 2.4 Mar‘13 2.6 Jan ‘14 Hash-based sharding Text Search V8 JavaScript engine Faster counts Aggregation Framework Multi-Data Center Deployments Improved Performance and Concurrency Need to edit this slide to make 2.6 look like the big deal. Also need a cooler intro!

What’s New Query System Improvements Integrated Text Search Enterprise-Grade Security Non-Stop Operations

Query System Improvements

Query System Improvments Index Intersection Pipelined Data Transformations New Update Operators New Set Operators

Index Intersection Simpler ad-hoc queries Existing Indexes can be combined to optimize a query Less Index Maintenance Smaller Working Set Lower Write Overhead More Adaptive

Pipelined Data Transformations $out for Aggregation Write results sets to a named collection No limit on output size Cursors back for results

New Update Operators $mul $min $max $currentDate New modifiers for $push Look at Scott H slides for update

New Set Operators $project operator in the aggregation pipeline now supports an expanded set of expressions $setIsSubset $setEquals $setDifference $setIntersection $setUnion $allElementsTrue $anyElementTrue

Redaction Field-level Security Access policies can be set in Query Language Two criteria to set Permissions Document Content User Privilege

Integrated Search

Text Search Now production-ready Integrated with MongoDB’s Query Engine Multi-language document support Text-search expressions can be used in the Aggregation Framework (not true in 2.5.4) Having text search available in MongoDB avoids the need to provision an external search cluster and replicate data to it for common search functions. Text search allows users to run text search queries in-place over MongoDB data, reducing cost and operational overhead to deliver search functionality.

Possibilities with Text Search Relevance ranking Boolean operators Language-specific tokenization and stemming Fielded search Field-weighted scoring Stop words Type-aware indexes Take slides from Text Search tech talk

Security

Security Authentication with LDAP (Enterprise only) x.509 Certificates User defined roles Auditing (Enterprise only) Windows Kerberos Support (And Windows Enterprise)

State of Security in MongoDB Authentication Who are you? X.509 authentication and Kerberos Authorization What can you do? User Defined Roles, Collection-level Access Control Auditing What have you done? DDL, User Manipulation, Authorization failure Things that are missing—queries, crud, access granted in general We don’t audit every single operation.

Operational Improvements

Improving Performance and Scalability Query Router Connection Pooling Bulk Write Operations Resource Overload Protection with $maxTimeMS Query Router Connection pooling: Connection pools for the MongoDB query routers can now be re-used by multiple MongoDB servers. This reduces the total number of connections needed in highly concurrent systems, lowering resource consumption by up to 5x. As a result, users can scale out to larger clusters supporting higher user load with less hardware. Bulk Write Operations: Write operations can now be applied in bulk to a MongoDB collection, increasing performance and enhancing scalability when ingesting data from high velocity data streams.   Users can configure the batch operation to either: Complete, reporting any write errors for later retries Stop the operation immediately if an error is returned. Resource Overload Protection: The new $maxTimeMS option (add link to docs) enables administrators to abort MongoDB queries and commands automatically if they exceed a pre-configured maximum time limit, with error messages returned to the client. Administrators can coordinate time-outs between the application, the driver and the database to prevent resource overload and ensure predictable performance. Time-outs can also be configured per query, affording high user flexibility.

Operational Improvements Background Secondary Indexing Mixed SSL Connections Expanded SNMP Support Background Secondary Indexing: Index builds running in the foreground of a MongoDB instance will block other database operations. To provide continuous availability, MongoDB therefore supports the creation of indexes in the background when they are running on the primary member of a replica set. MongoDB 2.6 extends this capability to secondary members of the replica set, which now replicate and perform the index build in the background as well. As a result, both primary and secondary replicas remain fully online, without operators having to manage rolling restarts. In addition, index creation operations that are interrupted due to a failure can now automatically resume when the node recovers. Mixed SSL Connections: Users now have the flexibility to run SSL and non-SSL connections to the same MongoDB server. As a result, administrators can avoid MongoDB downtime as they implement SSL across the cluster. Expanded SNMP Support: MongoDB Enterprise offers SNMP support, which can be used to integrate MongoDB with external monitoring solutions. Using MongoDB 2.6, Enterprise customers now have access to nearly the full range of metrics provided by the db.serverStatus() command, providing more detailed server metrics.

#MongoDB Thanks! Massimo Brignoli massimo@mongodb.com