Hotfix installation

Please make sure you consult the page Hotfix Installation Notes for any additional actions needed to complete the installation.

Main correction(s) provided

• CSV export handles missing properties on document for Vocabulary metadata.
• MongoDB query maxTime is fixed.
• The operations RenderDocument & RenderDocumentFeed are usable with REST.
• Unpublishing a source document loads only the published documents.
• The Signature Algorithm for SAML is configurable.
• The bulk status is kept for 24h in case of error.
• Security policies are taken into account by Annotations in NEV
• Files shared among Google Drive accounts are correctly uploaded with LiveConnect
• Aspera connector retrieves blob from Folderish File documents.
• CAS authentication redirects with a HTTP 302 when anonymous is enabled.
• Transaction timeout is prevented during updateThumbnailListener execution.
• A timeout is added to command line execution.
• CommandLine timeout is disabled on Windows.
• Picture migration is removed from startup.
• The local configuration for subtypes is inherited to subfolders.
• Carriage return characters are escaped in the CSV export.
• The returned HTTP status code of a REST/Automation request is based on thrown exception.
• The thumbnail is correctly generated for Audio documents.
• File content is unchanged after an exiftool failure.
• Docker image now uses LibreOffice 7.1.1
• Enricher priority is now taken into account when overriding an enricher.
• [Docker image] Fix package installation at runtime with packages/store mounted as a volume
• Possible NPE in KafkaLogTailer.close
• No principal with automation in the context of a scheduler event
• Fix CSV Import tests on huge files taking too long
• Update Aspera preview configuration with the new S3 Blob Provider
• WriteColdStorage permission not visible in permissions UI
• Fix cleanup of packages backup directory when using install-packages.sh
• Kafka topic not available after creation

New Feature(s)

• The maximum query execution time on MongoDB side is now aligned with the Nuxeo transaction.
• The new blob key is used for old blob used concurrently with async digest computation.
• The new field ecm:blobKeys is added to denormalize the list of all blob keys.
• A new property is added to compute blob digest asynchronously when Nuxeo doesn't see the blob content at upload time.
• Add new default picker page provider

Task(s)

• Improve pipeline error handling
• Use helmfile to ease a full nuxeo deployment with the nuxeo Helm chart
• CI Pipelines to build and promote LTS 2021 HF
• Upgrade Tomcat to 9.0.45
• Disambiguate unit test and preview K8s namespaces
• Fix random ConfigurationMarshallerTest.testPersistNuxeoConf*

Improvement(s)

• The user change token is used as WOPI item version.
• Additional logs are produced for S3 downloads.
• Workflow model import uses a cluster-wide lock.
• Nuxeo health check traces failures.
• The RemoteIp Valve is added to Tomcat configuration.
• The Bulk scroller completes a command in error when a query times out.
• S3 configuration parameters have been externalized to nuxeo.conf.
• Elasticsearch passthrough is configurable with certificates from private certificate authorities.
• Write a generic message for exception mapped to HTTP error code >= 500
• Improve filtering of timer metrics for Datadog
• Allow adding new templates without overriding the whole nuxeo.templates property
• Split functional test runs in tiers
• Deploy a released org.nuxeo:nuxeo-parent artifact for each build
• Tag the locally built LTS 2021 Docker image with 'latest-lts'
• Rely on the blob key for download using aspera
• Aspera add client timeout configuration
• Provide a way to detect that document is created or updated via CSV import in automation
• Make customisable Aspera Blob configuration

JIRA release notes

Click here to see the complete release notes