Query Workbench in Couchbase 4.5
Couchbase Server is the first and only database to combine the power of SQL with the flexibility, scalability, and performance of NoSQL. This release not only delivers even better performance – across querying, data access, and indexing – but also makes the transition from relational to NoSQL easier than ever, with features like query editing and auto schema discovery with the Query Workbench, high-speed backup/restore, and role-based access control.
POWERFUL QUERY TOOLS
The web-based admin console now includes Query Workbench – a built-in query editor with autocomplete for writing and executing queries as well as exploring the data model with automatically inferred schemas.
The query shell (CBQ) is now fully scriptable – create indexes, insert data, execute queries, and more with a script. In addition, when running Linux, commands can be piped to CBQ.
Administrators can now monitor active queries and connections as well as profile queries to identify and diagnose performance issues.
FASTER QUERYING & INDEXING
A new option enables memory-optimized data structures and persistent snapshots for indexes, improving indexing and query performance under write-intensive workloads.
The join syntax has been extended to support a new query optimization for filtering the right side first, with an index, resulting in added flexibility and improved query performance.
While N1QL (SQL for JSON) supports array processing within queries, it’s now possible to index specific fields of elements within nested arrays to improve query performance.
"At Seenit we store hundreds of gigabytes of machine learning data associated with our videos, including entities, visual tags, transcriptions, and sentiments. The improvements to array indexing in Couchbase Server 4.5 mean that we’re able to derive analytics and intelligence from these data points faster than ever before, and we’ve increased our analytics processing time by a factor of 5."
Dave Sterling, CTO, Seenit
A new consistency level, clients can ensure query results are as current as their last write - more consistency than eventual, better performance than strict.
A new option for secondary indexes enables the reuse of stale blocks and snapshots to limit the frequency of full compaction, improving query performance while reducing file size and disk IO.
DISTINCT queries are non-blocking, the IN predicate better utilizes secondary indexes, and the LIKE predicate can use secondary indexes via the SUFFIXES() function — and, pretty formatting can be disabled to improve response time.
FINE-GRAINED DATA ACCESS
The new sub-document API enables clients to add, remove, and modify individual fields (or read them), including objects and arrays, without reading and writing the entire document.
- Insert, update, and delete one or more fields
- Read one or more fields
- Check to see if a field exists
- Add or subtract from numeric fields
- Push elements to arrays (e.g. prepend / append / insert)
Administrators can now manage and rotate certificates from an independent authority for encrypted client-server and inter-cluster communication via SSL / TLS.
It’s now possible to grant different administrators different permissions by leveraging out-of-the-box roles for managing clustering, replication, indexing, and more.
IMPROVED DISASTER RECOVERY
A new backup and restore tool leverages streaming, incremental backups and an archive repository to support a faster recovery time with a closer recovery point.
DEVELOPER PREVIEW FEATURES
Built on the popular open-source project Bleve, support for term, fuzzy, phrase, match, and prefix queries has been added – faceting, scoring, and highlighting, too.
Red Hat OpenShift Enterprise 3
Red Hat OpenShift Enterprise 3 is in the process of being certified, with production support coming soon.