2021.4 (Released on August 25th, 2021)
Overview
The lasted version of Indexima – V2021.4 – is now available. This new version brings the following main new features:
Easy to use
- Indexima's new web console (available in preview since 2021.1) is ready for production use! It's now the default user experience when opening the console.
- Dictionaries access and type conversion improvements for accelerating queries on numerical types.
Easy to adopt
- Smart Analyzer: Add support for setting a table analyzer mode (defined in optimize_index.json) with an SQL command, to automate operations with many tables.
- Configuration simplification: Auto feed configuration parameters to simplify and ease Indexima setup and upgrade.
- External table management: Optimise source workload when delegating large requests to datasource warehouse.
Easy to implement
- High-availability enhancements (pre-version): new inter-node network stack, to support higher scalability scenario and failure resistance.
Useful links
- Download version 2021.4.
- Migrate from 2021.3 to 2021.4.
- Procedure for upgrading Indexima core engine (Galactica).
- Procedure for upgrading Developer Console (VisualDoop2).
End of life support policy
Please note that release 1.7.9 reached the end of life support as described by our end-of-life policy.
Detailed Release Note for the Indexima version 2021.4
Enhancements
- INDXMA-6603 (ZenD #1449): Add support for INSERT INTO SELECT with parenthesis/alias in CASE WHEN
- INDXMA-6583: Improve memory consumption when resizing dictionaries
- INDXMA-6514: Allows to import a schema under a new name
- INDXMA-6626: Support of views in SHOW TABLECOUNT
- INDXMA-6625: Improve support of views in API services
- INDXMA-6600: Extended support for stream INSERT from a SELECT operation
- INDXMA-6584: Improve memory garbage collection for sub-requests
- INDXMA-6573: Reduce memory usage in result output chunks
- INDXMA-6562: Improve error handling for join sub-requests
- INDXMA-6552: Optimise source workload when delegating large request
- INDXMA-6550: Optimize analyzer plan storing
- INDXMA-6480: Optimize memory allocations when node join a cluster
- INDXMA-6298 (ZenD #1355): Extend rights on default schema
- INDXMA-6451: Add SQL command to setup analyzer mode for a table
- INDXMA-6411: Support boolean as a native type
- INDXMA-5656: Improve log message upon index creation
Configuration
- INDXMA-6544: Fix migration.lock default path in the warehouse
- INDXMA-6650: Update template default value for warehouse
- INDXMA-6598 (ZenD #1408): Simplify Zookeeper connection string in the new interface
- INDXMA-6565: Enhance control for overflow of configuration parameters
- INDXMA-6438: Include galactica_ext.conf (dynamic parameters) in the diagnosis
Security Enhancements
- INDXMA-6450: Prevent usage of SHOW CONFIG for non-admin users
- INDXMA-6445: Hide password from URL in error messages of JDBC load assistant
- INDXMA-6667: Support encryption of connection passwords
- INDXMA-6661 (ZenD #1267): Extend support for HSTS header
- INDXMA-6120 (ZenD #1265): Improve protection against Cross-Site Request Forgery
External Table
- INDXMA-6496: Improve DATE_FORMAT delegation
- INDXMA-6484: Prevent returning partial results when killing a delegated query during fetch
- INDXMA-6424: Extend support of 'create table' wizard for Impala source
- INDXMA-6403: Improve snowflake synchronization in case of truncate
- INDXMA-6706: Add support for dynamic setup of Cognos parameter
- INDXMA-6602: Extend supported_db.json with Netezza
- INDXMA-6595: Cast boolean as an integer in aggregates on external tables
- INDXMA-6489 (ZenD #1402): Handle Boolean on Hive & impala external table
- INDXMA-6421: Add better support for Hive external table
- INDXMA-5618 (ZenD #1419): Improve closing delegated request upon request kill
Smart Table
- INDXMA-6691: Fix non-optimal selection of hyperindexes if a column appears several times in the SELECT statement
- INDXMA-6662: Fix edge case of SmartIndex infinite wait for add index
- INDXMA-6502 (ZenD #1403): Fix join key incorrect usage during query flattening
- INDXMA-6624: Extend support for complex prejoin aliases
- INDXMA-6606: Fix smart analysis not triggered on an external table
- INDXMA-6577: Fix SnowflakeSQLException upon index creation
- INDXMA-6574: Fix unknown column error upon index creation
- INDXMA-6542: Increase log detail of analyzer
- INDXMA-6535 (ZenD #1416): Fix wrong display of nb days in analyzer traces
- INDXMA-6490: Fix missing default smart optimizer mode for ingested tables
- INDXMA-5861: Invalidate sampling and cardinality store upon changing sampling size
Console
- INDXMA-6715: Fix Main Index flag in the console
- INDXMA-6592: Fix missing default index display in the console
- INDXMA-6587: Enhance display of metadata for views
- INDXMA-6546: Invalidate connection upon username modification
- INDXMA-6700: Extend error message for wrong API secret key
- INDXMA-6652: Allow changing cluster API secret key in the console
- INDXMA-6066: Fix incorrect query splitting when a query ends with an escaped backslash
- INDXMA-6507: Improve display of boolean types in the console
- INDXMA-6617 (ZenD #1296): Simplify the memory size reporting in the console
Bugs
- INDXMA-6674: Fix migration issue with 2 projects with the same name but different letter case
- INDXMA-6672: Fix show create for index with IN clause
- INDXMA-6671: Fix a case of out of bounds exception on add index with count distinct
- INDXMA-6668: Fix a join computation failure during load
- INDXMA-6664: Fix error migrating corrupted indexima projects
- INDXMA-6621: Fix jar dependency conflict on hadoop3
- INDXMA-6611: Fix closing file descriptor for table metadata
- INDXMA-6567: Improve error message Remote peer is closed
- INDXMA-6528: Fix error when deleting with timestamp selection
- INDXMA-6527: Improve queries history log flushing
- INDXMA-6520: Use connection identifier instead of name for Hive connection pool
- INDXMA-6511 (ZenD #1417): Add missing clearing of kill query manager
- INDXMA-6506 (ZenD #1407 / #1445): Save metadata after a compact operation
- INDXMA-6488: Fix offset for WeekofYear
- INDXMA-6487 (ZenD #1399): Fix DATE_FORMAT with the "day of the week" u flag
- INDXMA-6479: Fix protobuf dependency in Hadoop3 installer
- INDXMA-6478: Fix wrong query splitting for queries with a linebreak
- INDXMA-6454 (ZenD #1388): Improve error handling when sending twice the same index creation
- INDXMA-6402 (ZenD #1345): Fix denied ranger authorization when adding an index
- INDXMA-6369 (ZenD #1333): Fix diagnostic download from secondary masters
- INDXMA-6348: Fix edge case of table ambiguity on create table
- INDXMA-6556: Fix edge case when killing request after service restart