Version 23.2¶
23.2.8¶
2023-12-06
Improvements¶
- Use entity type ordering when indexing (#224877)
- Mark entity for indexing instead of indexing it directly (#224876)
Bug fixes¶
- DELETE_JOB fails on gcp storage (#224321)
- VidiCore Storage Priority is not respected for ThumbnailJob (#223359)
- CORS headers are not set on response when requesting file (#198119)
23.2.7¶
2023-11-15
Improvements¶
- Configuration of non-ascii character support in URI paths (#221075)
Bug fixes¶
- first query parameter is not ignored when cursor query parameter is used when listing files (#224766)
- VSA logging sometimes doesn’t work (#224662)
- Facets in collection search call are duplicated (#224024)
23.2.6¶
2023-10-19
Bug fixes¶
- Uploading a new essence version using chunked upload fails (#221828)
- Fail to export big file to FTP server (#197497)
- Auto-complete search request containing upper case letters doesn’t provide desired results (#195171)
23.2.5¶
2023-09-14
Improvements¶
- Add support for Oracle S3 (#222160)
Bug fixes¶
- S3FileWorker always reading object tags causes performance issues (#221997)
- Issues with file sequences (#219293)
- VidiCore should prevent a collection being its own child or parent (#218440)
23.2.4¶
2023-08-23
New features¶
- Add support for stored credentials for VSA (#214486)
Improvements¶
- Partial restore using file copy endpoint (#216689)
- Increase transfer block size (#220730)
- Parameter to set OpenSearch field mapping on init (#219539)
Bug fixes¶
- Wrong access level for GET thumbnail/spritesheet (#221446)
- Metadata Dataset to consider Nested Metadata Field Groups (#221244)
- Error from Solr when trying to index metadata field group without timestamp/created (#221230)
- ElasticSearch/OpenSearch status is wrong in self-test results (#220906)
- S3 object tagging errors are not ignored for Thumbnail storages (#220853)
- Revision is not updated when onlyReturnChanges is true (#220056)
- Accept Header being randomly selected by VidiCore when unset (#220008)
- Creating metadata-field with system=false causes NPE (#219291)
- Incorrect timecode for SCC export file (#219027)
- VSA logs too much (#218841)
- EXPORT job stuck in READY state (#218615)
- IMF import from S3 storage returns 403 error when using IAM roles (#215129)
- Keyword ‘NOW’ in the SearchDocument throws too many WARNING messages (#209005)
23.2.3¶
2023-06-22
Bug fixes¶
- Issue with Infinispan AdvancedCached computeIfAbsent in InMemorySchema (#219445)
23.2.2¶
2023-06-13
Improvements¶
- Detect circular locks and remove unnecessary locking (#219098)
Bug fixes¶
- S3 transfers stuck on import jobs (#219091)
23.2.1¶
2023-06-02
Bug fixes¶
- Cannot override openSearch path from configuration (#218845)
- Subpath not created in conjucntion with filename script on shape move endpoint (#218828)
23.2¶
2023-05-31
Read and write AWS S3 object tags via the Vidicore API¶
Files residing in a S3-bucket can now be tagged via the Vidicore API by setting key-value metadata on the file entity in Vidicore. This is useful for triggering AWS lifecycle policies based on criterias from Vidicore so that file objects can be automatically moved to different storage tiers in AWS.
Vidicore is also able to read tags from AWS and these will be stored in the same way as key-value metadata on the file entity.
For more information, see this section
Redundacy configuration for thumbnails¶
It is possible to add several paths into one thumbnail resource. This is useful for the shared file system where different paths are mapped to the same physical storage see vsa-storage. In this case, thumbnail resource is still accessible even if one of the paths is offline.
Note
When configuring thumbnail resources using json, the path element has been changed to accept an array of paths.
Heads up: Java 17 in Vidicore 23.3¶
For the next release, planned for end of August 2023, Vidicore will move to Java 17. For customers running the provided Debian and RedHat packages for on-prem or byoc deployment, be aware that the Java version needs to be updated.
New features¶
- Partition messages to ElasticSearch/OpenSearch (#214926)
- Support thumbnail resource redundancy configuration (#197480)
- Ability to transfer all files within a shape as one job (#214704)
- Write s3 metadata via s3 fileworker (#214323)
- Store sidecar subtitles as subtitleComponent on import (#203305)
- Mask credentials in the job log (#202706)
- Read s3 metadata via s3 fileworker (#201533)
Improvements¶
- Add possibility to inject and use custom job plugins (#218605)
- Documentation: _configuration_read role required to read confiuration properties (#215949)
- Make highlighting available in Collection Search Result (#200875)
- Include a parameter subpath in shape move endpoint (#218429)
- Update various packages according to Mend scan (#217931)
- Faster XML/JSON generation for large documents (#217899)
- Use _source filtering for inner_hits (#217898)
- Return all license properties via /API/version (#217003)
- Upgrade to infinispan 13 (#216539)
- Documentation: Default values for some configuration properties are not being displayed correctly in documentation (#215871)
- Add AWS SDK v2 to VidiCore (#215054)
- Substitute aliases with secrets in vidinet jobs (#215000)
- Bundle rabbitMQ with VidiCore (#214975)
- Export locations should support key-value-metadata (#214835)
- Update storageClasses for S3 (#214767)
- Remove erroneous error logging from Mp3Detector (#214588)
- Add support for setting stsEndpoint for s3 storage method (#213253)
- Return transient metadata fields when an /item/list?field= API is called (#207895)
- Documentation: Warning about using search highlight without specifying any highlight fields (#203507)
- Update to use official Opensearch client (#202647)
- Add AWS SDK v2 to vidispine-io (#201896)
Bug fixes¶
- Conform jobs with 60fps source material are not handled correctly (#218668)
- zero-padding is missing in get year (#218563)
- VidiCore should respect its timecode definition (#218544)
- Fix NPE in getting storage file (#217567)
- Notification for lost file doesn’t get triggered (#217435)
- Hashing should retry database operation (#217433)
- Create collection call with external-id that is already in use should return 409, not 400 (#217163)
- Document reindex not working with OpenSearch (#216928)
- Metadata reindex gets NPE (#216892)
- Set correct file state for S3 object that has been archived when storage method is vxa (#216815)
- IMF import fails if assetmap or other xml has byte order marks (#216623)
- Dataset metadata field values are missing in metadata notification (#215897)
- Incorrect JWK used to validate JWT token (#215490)
- Exception “No element has been retrieved” upon Reindexing a NoIndex Field (#215371)
- Metadata group indexing goes to wrong ES/OS workers (#215329)
- VSA not starting due to missing version information (#215291)
- Cannot update metadata group by remove and add mode (#214920)
- Cannot rename or delete files on sftp storage (#214810)
- Attempt to start library list job with empty field parameter should fail (#214317)
- MetadataFieldGroups in Collections can not be searched on, with opensearch (#211293)
- Show correct message while creating Poster resource (#210504)
- S3 to S3 transfer doesn’t work with Export job (#203172)
- Delete from a collection via PUT /collection/<collectionId> removes wrong entry (#200151)
Transcoder fixes¶
- Add subtitleComponents when doing shape deduction
- Make it possible to do a shape deduction on subtitle files
- Add support for autorotate for images