Release Notes¶
GraphDB release notes provide information about the features and improvements in each release, as well as various bug fixes. GraphDB’s versioning scheme is based on semantic versioning. The full version is composed of three components:
major.minor.patch
e.g., 9.11.2 where the major version is 9, the minor version is 11 and the patch version is 2.
Occasional versions may include a modifier after a hyphen, e.g., 10.0.0-RC1 to signal additional information, e.g.,
a test release (TR1, TR2 and so on), a release candidate (RC1, RC2 and so on), a milestone release (M1, M2 and so on),
or other relevant information.
Note
Releases with the same major and minor versions do not contain any new features. Releases with different patch versions contain fixes for bugs discovered since the previous minor. New or significantly changed features are released with a higher major or minor version.
GraphDB 10 includes the following components with their version numbers:
RDF4J
GraphDB Connectors
GraphDB Workbench
Their versions use the same semantic versioning scheme as the whole product, and their values are provided only as a reference.
GraphDB 10.4.2¶
Released: 20 November 2023
Component versions¶
RDF4J |
Connectors |
Workbench |
4.3.8 |
16.2.2 |
2.4.2 |
Important
GraphDB 10.4.2 contains multiple important bug fixes, including issues with SHACL, backups in cluster and a regression in SPARQL COALESCE.
Various third-party libraries were updated to address vulnerabilities.
We recommend everyone to upgrade.
GraphDB Engine & Cluster¶
New features and improvements¶
GDB-8933 Improved low disk space checks when importing data in parallel mode
Bug fixing¶
GDB-9078 Regress in handling SHACL shapes when migrating from GraphDB 10.2 or older
GDB-9032 GraphDB takes up a big chunk of off-heap memory when propagating backup to followers on slower networks
GDB-9010 SPARQL COALESCE regression leads to wrong results
GDB-8904 Incorrect error message in the log file during data import with low disk space
GDB-7342 Requesting “reinfer” fails to clear the previously inferred statements
GraphDB Workbench¶
Bug fixing¶
GDB-9132 “Talk to Your Graph” user interface does not work when the Workbench URL uses a context different from “/”
GDB-9117 Workbench SQL table preview return no results
GDB-9092 Toast messages not handling <>& symbols when reporting invalid IRIs
GDB-9074 When only one query or update is running it is not shown in the global monitoring status area
GDB-9026 Visual graph does not show result when the ignore predicates field is empty
GDB-9019 Workbench RDF Rank view confuses prefixed/unprefixed IRIs in some cases, which may prevent deleting a filter IRI
GDB-8336 Editing literals ending with a double quote is broken in Workbench Edit resource view
GraphDB 10.4.1¶
Released: 1 November 2023
Component versions¶
RDF4J |
Connectors |
Workbench |
4.3.6 |
16.2.1 |
2.4.1 |
Important
GraphDB 10.4.1 contains multiple important bug fixes, including issues with S3 backup and an issue related to statements not being processed correctly when using replace graph.
Various third-party libraries were updated to address vulnerabilities, including a patch for Spring’s vulnerability CVE-2016-1000027, even though GraphDB was not affected by it.
We recommend everyone to upgrade.
GraphDB Engine & Cluster¶
New features and improvements¶
GDB-9018 Regions should not be mandatory for S3 Backup and Restore
Bug fixing¶
GDB-8973 Literal index error on shutdown
GDB-8972 Unnecessary wrapping of exceptions may lead to loss of error message
GDB-8970 “Index is not writable” error when using preload with some datasets
GDB-8955 Cluster node hangs when receiving streaming entry after multiple unsuccessful snapshot attempts
GDB-8940 NPE when creating backup stored in S3
GDB-8935 External proxy’s node registration does not work with self signed certificates
GDB-8921 Remote locations cannot communicate through HTTPS with self signed certificates
GDB-8917 Lost statements during transactional update with replace graph in some cases
GDB-8857 Disable renaming a repository in a cluster setup
GDB-8794 Structures monitor endpoint does not return data in some cases
GDB-8768 Cannot add new node if an update is accepted right after the node add request
GDB-8664 Make the S3 client work with external bucketUri exposed over HTTPS
GDB-8470 Calling
gpt:listmay fail to process the result as a list in some casesGDB-7364 Cluster setup does not work with self signed certificates as it did in 9.x
GraphDB Workbench¶
New features and improvements¶
GDB-8988 Disable the edit button for renaming a repository in a cluster setup
GDB-8810 Add autocomplete subject, predicate, object, context fields in the rule edit form of ACL management
Bug fixing¶
GDB-9016 Global monitoring in top bar not working with cluster proxy
GDB-8986 Broken SVG export of class hierarchy and relationships
GDB-8910 Wrong color on info icon in “SQL table configuration” and “Edit SPARQL Template” views
GDB-8891 Talk to your graph: Sending a question, while another question is being processed will lead to unwanted behavior
GraphDB 10.4.0¶
Released: 17 October 2023
Component versions¶
RDF4J |
Connectors |
Workbench |
4.3.6 |
16.2.0 |
2.4.0 |
Important
GraphDB 10.4 introduces several significant features and improvements. Named graph and quad-based access control enables fine-grained security, providing flexibility through Access Control Lists (ACLs) and custom roles for authorizing data access below the repository level. A new chatbot and an integration with the ChatGPT Retrieval Plugin enables natural language queries about data stored in GraphDB. Improvements in system monitoring bring awareness of actions like backups and cluster recoveries, ensuring a more transparent user experience. Backup compression optimizes storage efficiency and performance and reduces both the time and costs needed to maintain backups.
Other features and improvements include better S3 support for cloud backups, adhering to consistent UX guidelines, better transaction traceability in the History plugin, an AWS operational guide, automatic transaction log size management, and updated third-party libraries for security and reliability.
- Named graph and quad-based security
The named graph and quad-based security feature in GraphDB is a more intricate access control system enabling precise permission settings beyond just the repository level, extending to the named graph level or the quad level which encapsulates an RDF statement. This feature introduces Access Control Lists (ACLs) that can be used to specify permissions based on any aspect of a quad — subject, predicate, object, or context. The ACLs support both permissive and restrictive policies, allowing the grant or denial of access to specific data as needed. Additionally, the feature provides flexibility to define custom roles, aligning permissions with users’ roles within the organization, ensuring that only authorized personnel can access sensitive data.
- Talk to Your Graph
The Talk to Your Graph feature in GraphDB integrates the ChatGPT Retrieval Plugin with GraphDB, aiming to boost ChatGPT’s question-answering capabilities regarding data stored in GraphDB. Users can access a chatbot in Workbench where they can ask natural language questions about their data, much like using the stock ChatGPT interface.
The integration provides a ChatGPT Retrieval Connector whose role is to synchronize RDF data with a vector database like Weaviate or Pinecone by converting RDF data into text documents, and then indexing it as vectors via the ChatGPT Retrieval Plugin.
A bonus use case allows for similarity search, where the new connector provides text query search and retrieves matching document IRIs, which can then be used for further joins in SPARQL, even without utilizing ChatGPT.
- System health awareness
System health awareness provides users with insights into actions like active backup, restore, and cluster recovery operations that were not easily visible. This helps to mitigate issues caused by users restarting GraphDB due to a lack of understanding. By enhancing various monitoring facilities in Workbench to include backups, restores and cluster recoveries, users will be better equipped to assess the system’s writable state, observe cluster node recovery processes, and understand when a node is engaging in backup creation or restoration. This feature will enrich the repository health check endpoint to reflect the writability, ensuring a more transparent and informative user experience.
- Backup compression
Backup compression is a transparent mechanism that reduces the size of GraphDB backups. This enhances the performance and storage efficiency of GraphDB through smaller backup sizes, which in turn lead to improved speed, especially when operating with slower storage systems. This feature also significantly reduces storage requirements for retaining multiple backups, resulting in cost savings and optimized resource allocation.
- Improved S3 support
Improved S3 support enhances cloud backup security and optimizes backup performance. Firstly, it intends to shift away from the current method of providing AWS credentials via query parameters to utilizing instance profile credentials for S3, thereby mitigating risks associated with static access ID and key, and reducing the need for secure storage services. Secondly, the feature focuses on optimizing S3 backup performance by leveraging parallel I/O operations of the Amazon client to expedite the backup process. This acceleration in backup times is crucial for supporting clients with larger datasets and facilitating GraphDB’s deployment on AWS, providing a robust and efficient S3 support mechanism within GraphDB.
- Consistent UX guidelines
The implementation of consistent UX guidelines addresses visual challenges related to contrast discernment and color blindness, enhancing the user interface’s accessibility. Adhering to the WCAG guidelines for contrast levels, efforts are channeled towards ensuring optimal contrast and accommodating color blindness while staying true to the original color scheme. This feature goes further by also addressing the inconsistencies in button positioning throughout Workbench. By aligning with common practices adopted by notable platforms like Apple and Google for button arrangements, we provide a more intuitive and user-friendly interface.
- Distinguish History Plugin transactions
The Distinguish History Plugin transactions feature in GraphDB enhances transaction traceability and analysis by assigning stable IDs to each transaction, replacing the current timestamp-only identification. This enhancement simplifies referencing of specific transactions and understanding their impact over time. This feature also enables tracking and querying of the user behind each transaction, facilitating auditing processes and identification of potential training or guidance needs. By implementing these improvements, users can navigate and analyze the transaction history more efficiently, promoting informed decision making and a comprehensive understanding of data evolution within GraphDB.
- AWS operational guide
The operational guide for GraphDB on AWS facilitates seamless GraphDB cluster deployments on AWS, anticipating a significant shift towards cloud platforms in 2023 and 2024. This guide targets DevOps professionals, covering three deployment scenarios: client-operated GraphDB deployment on AWS with Ontotext’s third-level support, Ontotext-operated managed service on AWS, and client-operated managed service acquired through the AWS Marketplace. By standardizing deployment architectures and options, this initiative seeks to minimize the setup effort and time for teams adept with AWS but unfamiliar with GraphDB, thus expediting the adoption and operational readiness of GraphDB on AWS.
- Automatic transaction log size management
Automatic transaction log size management reintroduces a crucial feature from the GraphDB 9 cluster. This feature ensures an optimal transaction log size by auto-truncating old entries while retaining essential information for cluster synchronization and recovery, enhancing system performance and reducing storage needs.
- Stay up-to-date with the latest versions of third-party libraries
As a general strategy to offer a secure and reliable product, we strive to provide up-to-date versions of third-party libraries. This includes both features and bug fixes provided by the libraries and also addresses newly identified public vulnerabilities.
GraphDB Engine & Cluster¶
New features and improvements¶
GDB-8809 Add support for AT+JWT token type in OpenID tokens
GDB-8737 Expand the healthcheck API to report the writable status of a repository
GDB-8648 As a DB admin, i need quad-based access control for fine-tuned control of access to repository data
GDB-8401 As a DB admin, I want S3 backups to support instance profile credentials
GDB-8373 As a DB admin, I want S3 backups of GraphDB to leverage parallel I/O operations
GDB-7897 As a DB admin, I want GraphDB to automatically truncate cluster transaction metadata
Bug fixing¶
GDB-8922 External proxy is unable to record the latest leader
GDB-8909 Unclean shutdown leads to issues with the entity pool
GDB-8887 Concurrent modification exception during external proxy communication
GDB-8815 Check for enough disk space does not take into account space already taken by the repository
GDB-8811 Cluster load balancer does not balance fairly between all nodes
GDB-8803 Unable to restart GraphDB after reaching fatal low disk space threshold
GDB-8797 GPT client in GraphDB does not use HTTP proxy set via the Java instance
GDB-8738 Invalid handling of bad restore request in cluster
GDB-8736 Follower can’t create similarity index due to cannot directly process transactions restriction
GDB-8638 S3 backups time out for seemingly no reason
GDB-8619 In version 10.3.x the binding with UNION trough similarity-plugin does not evaluate correctly
GDB-8571 Cluster node cannot properly shut down during a backup procedure
GDB-8559 Deleting the cluster during a backup via a follower node leads to errors and broken cluster status
GDB-8508 Import data via importrdf preload does not check for low disk space
GDB-8507 Importrdf tool in parallel mode may throw a NPE when the disk space is low
GDB-8451 When a cluster is scaled down the external proxy keeps pinging a node address if it was predefined in its configuration
GraphDB Workbench¶
New features and improvements¶
GDB-8770 As a user, I want to ask questions that ChatGPT may not have information about and receive better responses by leveraging the enriched knowledge stored in GraphDB
GDB-8541 As a user, I expect the buttons to be positioned in a generally accepted way and consistently throughout the workbench
GDB-8540 As a visually challenged user, I need standards-compliant contrast and color scheme in the workbench
GDB-8281 ND-JSONLD file format missing when downloading data in Workbench
GDB-8280 ND-JSONLD and some alternative extensions for other formats missing when importing data in Workbench
GDB-7847 As a user, I want to track running backup/cluster-recovery/restore operations
Bug fixing¶
GDB-8730 Improved usability when adding a namespace in Workbench
GDB-8644 Improved handling of logout/relogin with a different user when the new user has no access to the previously selected repository
GDB-8524 Improved error message in Workbench import when running out of disk space
GDB-8429 Logout does not logout the user from all open browser tabs
GDB-8425 Cannot create/edit/delete user with ‘.’ in the username on Firefox with security
GDB-7782 Pasting a long query into the SPARQL editor causes an overflow in the line numbers
GDB-7116 Some icons in the status bar have on background, which causes issues when scrolling
GraphDB Connectors & Plugins¶
New features and improvements¶
GDB-8769 As a developer, I want to synchronize RDF data from GraphDB to a vector database
GDB-7777 As a user, I need a way to distinguish history plugin transactions and track the user that introduced a change
Bug fixing¶
GDB-8604 Wrong evaluation of some SPARQL SERVICE clauses with a connector query inside
GDB-8582 Elasticsearch connector error “IOException: entity content is too long for the configured buffer limit”