v4.2.0-1

Enhancements

• Updated to use Neo4j 4.2.0 as the new baseline (Release notes)
• Big thanks to @peter-c-larsson who added Azure backup support! (Ref)
• Big thanks to @jtyr - users may specify service accounts via values.yml (Related ticket)
• Big thanks to @maxbischoff - users may set an initial neo4j password with reference to a particular key in a kubernetes secret (PR)
• Certain best practice guidance has been added to documentation (Ref)

Fixes

• #114 resources incorrectly specified at the top level
• Documentation improvements from user feedback PR

v4.1.3-1

Enhancements

• Update to use latest Neo4j 4.1.3 as a baseline. (Related Neo4j 4.1.3 changelog)
• The backup/restore containers now work as a CronJob on a regular schedule, permit hot backups, and support Amazon S3 as a backup set target. #100 - these containers are not backwards compatible, so take due notice of changes in the documentation and required extra parameters (thanks to @srfaytkn)
• Use cgroup memory setting if available #111 (thanks to @eastlondoner)

Fixes

• Fixes #116 which dealt with an error in the documentation detailing how to disable APOC
• Fixes #119 (incorrect memory example for standalone) and related issue #113
• Fixed an error in helm template documentation (thanks to @achantavy ) eb48a1d
• Fixed broken documentation links dealing with external exposure (thanks to Neo4j field) #108
• Fixed documentation typos (thanks to @eNeRGy164 ) #109

Release build link here

Iterative improvements

Enhancements

• Allow arbitrary secrets and multiple configuration maps to be injected into the environment #102
• Label the cluster services that expose bolt/http #96
• Permit custom terminationGracePeriodSeconds #88

Bug Fixes

• Fixed error in core member resources specification #87

Other/Internal

• Make YAML more consistent #98
• Build chain improvements for documentation
• Mats contribution: CPUspecs #85

Flexibility Improvements

Improvements

• New documentation site! old documentation now points here.
• Layered configuration makes it easier to express certain scenarios #82
• Label selectors are overridable #81
• Allow any configuration to be overridden on a pod basis #79

Bugfixes

• Issue #67 hard-coded cluster domain #77

Bugfix release

• Fixes custom pod annotations and labels, and provides easy examples of how to use them #73
• Fixes additionalVolume & additionalVolumeMount specifications #72
• Adds a long termination grace period for Neo4j pods so that on shutdown, they can properly checkpoint and flush remaining page cache to disk before shutting down, helping to prevent unclean shutdowns. #71

Neo4j 4.1.1

• Upgrade default Neo4j provided to 4.1.1 #62
• Neo4j log file storage (debug.log, query.log) is now mapped to persistent storage by default #61
• Cluster discovery proceeds against the first 3 core members only #63
• Added documentation and examples on anti-affinity rules #64
• Fixed DNS caching error that under some circumstances could prevent cluster formation

4.1.0-3

• Support for custom pod labels and annotations #55
• Support for additional custom volumes and mounts #54
• Permit "full name override" for custom naming of kubernetes resources #53
• More flexible plugin configuration, including easy options to install Graph Data Science, and Neo4j Streams #51
• Better support for scaling core members (i.e. during rolling upgrade operations) #50
• Better user guide & documentation organization #43
• Bugfix: memory settings are properly passed down into the pods #44

Bugfix release

• Add correct PVC mapping for read replicas #39
• Update read replicas horizontal pod autoscaler reference #38
• Add documentation on horizontal pod autoscaling for read replicas
• Internal neo4j documentation feedback / improvements

Neo4j 4.1.0

• Neo4j 4.1.0 #31
• Converts the read replica set into a disk-backed statefulset #30
• Guidance on Fabric in Kubernetes #34
• Guidance on Rolling Updates in Kubernetes #33

Single Instance, backup, and restore

• Includes #23
• Includes #21
• Includes #22