Releases: orientechnologies/orientdb
Releases · orientechnologies/orientdb
3.2.33
This release include mostly fixes in the query engine, for both correctness and performance improvement
trough use of indexes in additional use cases, it also deprecate some APIs that will be removed in the
next major release.
Changes
Core
- Marked methods deprecated for removal in the next major
- Make sure to handle record with missing id (lightweight edges) in delete query, issue #10239
- Re-add support of indexes for between queries in new query engine
- Return every time the correct document instance in APis and queries, issue #10280
- Re-add fetch from variable for traverse in the new query engine implementation, issue #10281
- Avoid to use index when the condition is empty list
- Make sure to respect sorting on IN conditions with indexed fields
- Handle correctly lightweight edges in save
- Support of indexes in 'is null' operator
- Corrected ORestricted case for count operations in new query engine
Graph
- Make sure to bind g variable for groovy-gremlin scripts
Server
- Update the root password if env variable is set, issue #10273
Artifacts
orientdb-community-3.2.33.tar.gz
orientdb-community-3.2.33.zip
3.2.32
This released include a small set of fixes and the upgrade to the latest gremlin release,
some fixes have been pulled from downstream forks and are made by
@andrii0lomakin and @leostryuk, also a pull request was made by @timw with the Lucene fix.
Changes
Core
- Fix bug in SQL engine variable name resolution
- Keep order for asSet() function
- Minor optimization in query engine class/cluster filtering
- Make sure to load record from database all the times by query engine
- Make sure that toJson of list of ids get expanded, issue #10264
Lucene
- Fix IndexReader ref leak when Lucene index modified and read in transaction
Security Module
- Removed copied code in favor of dependency
OrientDB Gremlin Driver
- Update gremlin to the latest release 3.7.2
Artifacts
orientdb-community-3.2.32.tar.gz
orientdb-community-3.2.32.zip
Contributors
andrii0lomakin, timw, and leostryuk
Assets 2
3.2.31
Changes
Core
- Reduced logging for block size detecting
- Add support of empty array as from target of query
- Make sure to run let query every record when there is a context variable
- Make sure to be able to start and load databases on startup also when not run in distributed mode
Tools
- Add skipping of failed record on import, improved import logging add content of failed record
- Corrected parsing of rid to allow white space around it in import
- Corrected imported to avoid to terminate in case of failure to read a record
Distributed
- Make sure to retry to install the database later if it fail to install the first time
Artifacts
orientdb-community-3.2.31.tar.gz
orientdb-community-3.2.31.zip
3.2.30
Changes
Core
- Fix disk compatibility on some specific CPU architectures
Tool
- Add Support of preserving record ids on database import
Artifacts
orientdb-community-3.2.30.tar.gz
orientdb-community-3.2.30.zip
3.2.29
Changes
Core
- Fix issue on delegated records on raw access, issue #10171
- Minor fix in transaction not document record handling
- Make sure to check the security policy record exists before creating instance
Distributed
- Correct id swap revert in distributed transaction during retry
- Make sure to execute full sync in case of delta sync failure
Artifacts
orientdb-community-3.2.29.tar.gz
orientdb-community-3.2.29.zip
3.2.28
Changes
Core
- Check if the index exists before checking it's properties
- Make sure to initialize maps on update of value in query engine
- Add name to the OrientDB internal timer
- Removed optimization in query engine that cause failing to update linked records, issue #10077
- Corrected query with content of nested JSON documents to use embedded class
- Add count on the create link result set
- Make sure to call on drop on database drop
- Double Write Log is fsynced on writes. Looks like usage of SYNC flag is not enough
- Double Write Log is not truncated on file flush but only if it reaches file limit
- Double Write Log is truncated every time there are tail segments in queue
- We do not need to make fsync during Write Ahead Log truncation it will be forced by Double Write Log itself
- sync flag is removed for Double Write Log channel because we perform sync on data write
- Dirty counter is updated on write sync
Distributed
- Fixed database sync with enterprise agent
Etl
- Fix concurrent issues in not x86 architectures
Client
- removed duplicate cancel from client code
Artifacts
orientdb-community-3.2.28.tar.gz
orientdb-community-3.2.28.zip
3.2.27
Changes
Core
- Minor correction on value conversion for comparison in queries
- Corrected implementation of tracing executor to trace source properly
- Minor fix on storage open failure
- Correct set of target class when query target a cluster
Client
- Minor fix on retrying next host from client
Distributed
- make sure to ignore empty response from distributed requests
- Refactor distributed configuration management to be handled by distributed context
Server
- Minor fix on wait for shutdown
Artifacts
orientdb-community-3.2.27.tar.gz
orientdb-community-3.2.27.zip
3.2.26
Changes
Core
- Corrected order by with database locale
- Used the correct property name in alter property statement propagation
- Make sure that old query engine use public schema API for create property
- Re-enabled tests for bonsai tree with fixes
- Simplification of index implementation in storage and lucene
- Introduced executor for tracking submitter in case of exception
- Simplification and optimization in query engine results
- Corrected insert content query to respect default values
Distributed
- Report DDL query errors in distributed correctly
- Make sure the distributed index drop works in the same way of embedded
- Minor change in configuration reading on context initialization
Artifacts
orientdb-community-3.2.26.tar.gz
orientdb-community-3.2.26.zip
3.2.25
3.2.24
Changes
Core
- Optimized OResult to load records lazily
- Optimized OResult to avoid early load of link in properties
- Minor optimizations in the query engine runtime
- Fix insert into index step to use the right value
- Minor dependencies updates
- Fix regression in rid comparison in case of embedded query
- Fix regression in direct insert into index query
- Fix regression in update index step
- Avoid storage to go in read only mode if index key type validation fails