Getting started: Release notes
Changelog for Linkurious Enterprise v4.2.1
Release Date: 2025-06-18
New Features & Improvements (13)
- (v4.2.1) Deployment: new cluster mode for fault tolerance
- (v4.2.1) Entity Resolution: various user experience, performance and stability improvements
- (v4.2.1) Node grouping: node groups can be collapsed and expanded to simplify the visualization
- (v4.2.1) Configuration: Automatically migrate legacy Azure SSO OAuth URLs
- (v4.2.1) Documentation: explain the order in which styles rules are applied
- (v4.2.1) Security: Update Node.js to v22.15.1
- (v4.2.0) Security: Update to Node.js v22
- (v4.2.0) Embedded Elasticsearch: Update to Elasticsearch 8.17.4
- (v4.2.0) Memgraph: Add support for Memgraph 3.1
- (v4.2.0) Neo4j: Add support for Neo4j 2025.03
- (v4.2.0) Security: Enforce the cookie secret configuration to cryptographically secure
- (v4.2.0) User-data store: Add support for MariaDB 11.4
- (v4.2.0) User-data store: Add support for MySQL 8.4
Bug Fixes (50)
- (v4.2.1) Configuration-Migration Plugin: the migration fails for some node-grouping rules
- (v4.2.1) Data-Table plugin: the plugin crashes when the graph query fails with an 500 error
- (v4.2.1) Neo4j: failure to reconnect to a Neo4j cluster after the main Neo4j node goes offline
- (v4.2.1) Authentication: error when the optional tenantID field is omitted from the Azure SSO configuration
- (v4.2.1) Data-source: a graph database rebuild is not detected if the data-source is connected
- (v4.2.1) Deployment: sometimes, the server fails to start and the error is not logged
- (v4.2.1) Queries: no message in the query panel when there are no queries to display
- (v4.2.1) Queries: when a query returns more nodes than allowed, the truncated graph is displayed before the user dismisses the warning
- (v4.2.1) Visualization: in the multi-selection table, the currency format is not applied for properties with a whitespace in their name
- (v4.2.1) Visualization: when using the radial layout, removing the root node sometimes causes an error
- (v4.2.1) Workspace: setting the visualization text background to null prevents adding new edges
- (v4.2.0) Google Spanner: When the "keyFile" is a relative path, it does not point to the data folder
- (v4.2.0) Admin: In the groups management page, property-key access rights are shown as disabled when enabled
- (v4.2.0) Admin: Wrong error displayed when trying to edit the groups of an external user
- (v4.2.0) Alerts: When the data-source is disconnected while an alert is running, it does not stop gracefully
- (v4.2.0) Amazon Neptune: Error after editing a node that was just created
- (v4.2.0) Amazon Neptune: Indexing fails on Elasticsearch / OpenSearch
- (v4.2.0) Data-sources: Offline data-sources are no longer automatically detected
- (v4.2.0) Deployment: The process manager logs an error while shutting down
- (v4.2.0) Documentation: The Azure SSO configuration instructions are outdated
- (v4.2.0) Entity Resolution: "Socket hang up" error when sending requests to the entity resolution server
- (v4.2.0) General: Content with right-to-left format (e.g. arabic text) is not displayed correctly in some places
- (v4.2.0) Geo mode: Inconsistent behavior after removing all nodes in geo mode and switching back to node-link mode
- (v4.2.0) Geo mode: The layout is not recomputed after switching from geo-mode to node-link mode
- (v4.2.0) Image export: A progress bar appears when scrolling horizontally to the right in the modal
- (v4.2.0) Image export: Elements filtered by the timeline are not filtered in the exported image
- (v4.2.0) Image export: The image size is sometimes not consistent with the size of the preview
- (v4.2.0) Memgraph: Error when sampling schema while schema-cache and auto-indexes are enabled
- (v4.2.0) Neo4j: A data-source in "needs config" state leaks Bolt connections over time
- (v4.2.0) Neo4j: Credential mapping does not work with secure connections
- (v4.2.0) Neo4j: Graph schema sampling can timeout for large graphs
- (v4.2.0) Node Grouping: Sometimes, enabling node-grouping breaks the visualization
- (v4.2.0) Node Grouping: Sometimes, the graph layout becomes broken after un-grouping
- (v4.2.0) Node Grouping: The Layout is not recalculated after adding new nodes to node groups
- (v4.2.0) Node Grouping: Wrong placement of groups after removing and adding back same group
- (v4.2.0) Queries: Cypher autocomplete does not escape properties, edge-types or node-categories when they contain special characters
- (v4.2.0) Queries: Graph elements not extracted from collections on Neptune Cypher
- (v4.2.0) Queries: In the "delete query" modal, closing the modal with "x" deletes the query
- (v4.2.0) Queries: Path results are not correctly parsed on Amazon Neptune
- (v4.2.0) Queries: The query panel never recommends queries when the graph selection is empty
- (v4.2.0) Queries: The query panel suggests query templates that cannot be run with the current selection
- (v4.2.0) Search: An error is shown when clicking "Enable later" after connecting a data-source
- (v4.2.0) Search: If a node-category is missing from the graph schema, search can be broken
- (v4.2.0) Search: With Neo4j Search, fuzzy search does not work
- (v4.2.0) Security: The cookie secret is not randomly generated for each deployment
- (v4.2.0) Visualization: An error message is shown after removing a large amount of nodes (500+)
- (v4.2.0) Visualization: In the filter panel, numerical filters do not work (lower & upper bound inputs are always considered invalid)
- (v4.2.0) Visualization: Node and edge captions are not truncated in the selection panel
- (v4.2.0) Visualization: Searching for users in the sharing modal can show an empty result
- (v4.2.0) Visualization: The hierarchical layout sometimes places nodes in the wrong layer
Breaking changes (5)
- (v4.2.1) If you are using the "fault tolerance" setup documented in previous version of the product, there are no breaking changes but it is strongly recommended to use the new cluster mode instead. The cluster mode has better cross-instance consistency guarantees, and makes it easier to configure the primary and secondary server instance correctly. For details, see https://doc.linkurious.com/admin-manual/4.2/cluster-mode/
- (v4.2.1) If you are using Entity Resolution, you must update the Entity Resolution server to v1.1.0 and clear all records after updating to this version.
- (v4.2.0) Configuration files with secrets encoded before Linkurious Enterprise v4.0.25 can no longer be decrypted. To avoid any issue, Linkurious must be updated to v4.1 before updating to v4.2.0, so that secrets in configuration files are migrated.
- (v4.2.0) After this update, users using the embedded Elasticsearch must re-index their data-source.
- (v4.2.0) If your configuration contains a server.cookieSecret value that is shorter than 64 characters, you must replace it with a longer value.