Releases: scalar-labs/scalardb
v3.13.0
Summary
This release includes a lot of enhancements, improvements, bug fixes, and vulnerability fixes.
Community edition
Enhancements
- Added dynamic arbitrary filtering for non-JDBC databases. (#1682)
- Added the Insert, Upsert, and Update operations to the transactional API. (#1697)
- Added YugabyteDB adapter as one of JDBC storages (#1710)
- Added Group Commit feature for Coordinator Table (#1728)
- Allowed directly executing CRUD operations with transaction managers. (#1755)
- Added support for arbitrary filtering for partition scan and index scan. (#1763)
- Added a single CRUD operation transaction manager. This transaction manager implementation does not allow beginning a transaction by calling
begin()
/start()
. It only allows directly executing CRUD operations from the transaction manager. (#1793) - Added support for arbitrary filtering for get operations. (#1834)
- Added a new Admin API
admin.getNamespacesNames()
to list the user namespaces. Though, this API won't return a namespace that does not contain a table. From ScalarDB 4.0, we plan to improve the design to suppress this limitation. (#2002)
Improvements
- Removed the hard-coded collation for MySQL and SQL Server in the JDBC adapter. As a result, the collation configured in the underlying database will be used when creating tables. (#1518)
- Added error codes to the error messages of Schema Loader. (#1564)
- Performance improvement of the group commit by using priority queue in the background worker. (#1641)
- Refactored scan with filtering. (#1715)
- Avoided creating an internal unique index as much as possible to reduce resource consumption and improve performance. (#1723)
- Changed the hard-coded password for the Oracle user to a more secure one in the JDBC adapter. (#1765)
- Update base image of container image. This update fixes an OOM issue on a Kubernetes with cgroup v2 environment. In the previous versions, if you use a Kubernetes cluster with cgroup v2, you might face an OOM-killed issue. (#1826)
- Added capability to specify global properties for all storages in multi-storage. (#1486)
- In MySQL, ScalarDB
FLOAT
type is changed fromDOUBLE
toREAL
(single-precision floating-point value) (#2000)
Bug fixes
- Upgraded the base image to fix security issues. CVE-2023-47038 (#1522 #1521)
- Upgraded the PostgresSQL lib to fix security issues. CVE-2024-1597 (#1547)
- Fixed a bug where
NullPointerException
occurs during theEXTRA_READ
validation when scanning records in a transaction, but some of them are deleted by other transactions. (#1624) - Fixed a bug where lazy recovery was not executed for the implicit pre-read of put and delete operations. (#1681)
- Fixed a bug where users could see inconsistent results when scanning records by an index key after putting the related records in Consensus Commit transactions. (#1727)
- Upgraded
grpc_health_probe
to fix security issues. CVE-2024-24790, CVE-2023-45283, and CVE-2023-45288 (#1980) - Fixed snapshot management issues. (#1976)
- Fix a bug of the import-table feature that it could access tables in other namespace that have the same table name when using MySQL storage. (#2001)
Enterprise edition
Enhancements
ScalarDB Cluster
- Added support for the insert mode of the Put operation introduced #1679 in ScalarDB Cluster.
- Added support for insert, upsert, and update APIs introduced in #1697 in ScalarDB Cluster.
- Added support executing a CRUD operations in a one-shot transaction.
- Added support for arbitrary filtering for partition scan and index scan introduced in #1763 to ScalarDB Cluster.
- Added support for transaction managers other than Consensus Commit to ScalarDB Cluster.
- Added support for the single CRUD operation transaction manager introduced in #1793 in ScalarDB Cluster.
- Added support for arbitrary filtering for get operations introduced in #1834 to ScalarDB Cluster.
- Added support for
DistributedTransactionAdmin.getNamespaceNames()
ScalarDB SQL
- Added support for the single CRUD operation transaction manager introduced in #1793 to ScalarDB SQL.
- With this update, users now have several ways to access ScalarDB-managed namespaces in ScalarDB SQL.
Improvements
ScalarDB Cluster
- Added error codes to the error messages of the Auth module.
- Added error codes to the error messages.
- Added TLS support for the Prometheus exporter. With this change, when enabling TLS (setting
scalar.db.cluster.tls.enabled
totrue
) in ScalarDB cluster nodes, the Prometheus exporter also starts with TLS (HTTPS). - Update base image of container image. This update fixes an OOM issue on a Kubernetes with cgroup v2 environment. In the previous versions, if you use a Kubernetes cluster with cgroup v2, you might face an OOM-killed issue.
ScalarDB GraphQL
- Added error codes to the error messages.
- Update base image of container image. This update fixes an OOM issue on a Kubernetes with cgroup v2 environment. In the previous versions, if you use a Kubernetes cluster with cgroup v2, you might face an OOM-killed issue.
ScalarDB SQL
- Added error codes to the error messages.
- Changed the packages for
ConditionSetBuilder
andAndConditionSet
. - Allowed using the
EXISTS
keyword for theCREATE/DROP COORDINATOR TABLES
statements. - Update base image of container image. This update fixes an OOM issue on a Kubernetes with cgroup v2 environment. In the previous versions, if you use a Kubernetes cluster with cgroup v2, you might face an OOM-killed issue.
- Improved performance of selection queries with filtering by exploiting partition and index scans.
Bug fixes
ScalarDB Cluster
- Upgraded the base image to fix security issues. CVE-2023-47038
- Upgraded the Kubernetes Client Java lib to fix security issues: CVE-2024-25710 and CVE-2024-26308.
- Upgraded
grpc_health_probe
to fix security issues. CVE-2024-24790, CVE-2023-45283, and CVE-2023-45288 - Fixed a bug where incorrect results are returned when executing SELECT queries with the same column names.
ScalarDB GraphQL
- Upgraded the base image to fix security issues. CVE-2023-47038
ScalarDB SQL
- Upgraded the base image to fix security issues. CVE-2023-47038
- Fixes a bug that Spring Data JDBC for ScalarDB doesn't work with Spring Boot 3
- Fixed a bug where incorrect results are returned when executing SELECT queries with the same column names.
- Upgraded
grpc_health_probe
to fix security issues. CVE-2024-24790, CVE-2023-45283, and CVE-2023-45288
v3.12.3
Summary
This release includes several improvements, bug fixes, and vulnerability fixes.
Community edition
Improvements
- Refactored scan with filtering. (#1715)
- Changed the hard-coded password for the Oracle user to a more secure one in the JDBC adapter. (#1765)
- Update base image of container image. This update fixes an OOM issue on a Kubernetes with cgroup v2 environment. In the previous versions, if you use a Kubernetes cluster with cgroup v2, you might face an OOM-killed issue. (#1826)
Bug fixes
- Fixed a bug where
NullPointerException
occurs during theEXTRA_READ
validation when scanning records in a transaction, but some of them are deleted by other transactions. (#1624) - Fixed a bug where lazy recovery was not executed for the implicit pre-read of put and delete operations. (#1681)
- Fixed a bug where users could see inconsistent results when scanning records by an index key after putting the related records in Consensus Commit transactions. (#1727)
- Upgraded
grpc_health_probe
to fix security issues. CVE-2024-24790, CVE-2023-45283, and CVE-2023-45288 (#1980) - Fixed snapshot management issues. (#1976)
- Fix a bug of the import-table feature that it could access tables in other namespace that have the same table name when using MySQL storage. For example, in the following situation, the metadata of the columns
pk_unexpected
andcol_unexpected
ofns2.tbl1
are handled and the import-table feature fails due to unsupported data types. (#2001)
Enterprise edition
Improvements
ScalarDB Cluster
- Update base image of container image. This update fixes an OOM issue on a Kubernetes with cgroup v2 environment. In the previous versions, if you use a Kubernetes cluster with cgroup v2, you might face an OOM-killed issue.
ScalarDB GraphQL
- Update base image of container image. This update fixes an OOM issue on a Kubernetes with cgroup v2 environment. In the previous versions, if you use a Kubernetes cluster with cgroup v2, you might face an OOM-killed issue.
ScalarDB SQL
- Changed the packages for
ConditionSetBuilder
andAndConditionSet
. - Update base image of container image. This update fixes an OOM issue on a Kubernetes with cgroup v2 environment. In the previous versions, if you use a Kubernetes cluster with cgroup v2, you might face an OOM-killed issue.
Bug fixes
ScalarDB Cluster
- Upgraded
grpc_health_probe
to fix security issues. CVE-2024-24790, CVE-2023-45283, and CVE-2023-45288
ScalarDB SQL
- Fixed a bug where incorrect results are returned when executing SELECT queries with the same column names.
- Upgraded
grpc_health_probe
to fix security issues. CVE-2024-24790, CVE-2023-45283, and CVE-2023-45288
v3.11.3
Summary
This release includes several improvements, bug fixes, and vulnerability fixes.
Community edition
Improvements
- Refactored scan with filtering. (#1715)
- Changed the hard-coded password for the Oracle user to a more secure one in the JDBC adapter. (#1765)
- Update base image of container image. This update fixes an OOM issue on a Kubernetes with cgroup v2 environment. In the previous versions, if you use a Kubernetes cluster with cgroup v2, you might face an OOM-killed issue. (#1826)
Bug fixes
- Fixed a bug where
NullPointerException
occurs during theEXTRA_READ
validation when scanning records in a transaction, but some of them are deleted by other transactions. (#1624) - Fixed a bug where lazy recovery was not executed for the implicit pre-read of put and delete operations. (#1681)
- Fixed a bug where users could see inconsistent results when scanning records by an index key after putting the related records in Consensus Commit transactions. (#1727)
- Upgraded
grpc_health_probe
to fix security issues. CVE-2024-24790, CVE-2023-45283, and CVE-2023-45288 (#1980) - Fixed snapshot management issues. (#1976)
- Fix a bug of the import-table feature that it could access tables in other namespace that have the same table name when using MySQL storage. For example, in the following situation, the metadata of the columns
pk_unexpected
andcol_unexpected
ofns2.tbl1
are handled and the import-table feature fails due to unsupported data types. (#2001)
Enterprise edition
Improvements
ScalarDB Cluster
- Update base image of container image. This update fixes an OOM issue on a Kubernetes with cgroup v2 environment. In the previous versions, if you use a Kubernetes cluster with cgroup v2, you might face an OOM-killed issue.
ScalarDB GraphQL
- Update base image of container image. This update fixes an OOM issue on a Kubernetes with cgroup v2 environment. In the previous versions, if you use a Kubernetes cluster with cgroup v2, you might face an OOM-killed issue.
ScalarDB SQL
- Changed the packages for
ConditionSetBuilder
andAndConditionSet
. - Update base image of container image. This update fixes an OOM issue on a Kubernetes with cgroup v2 environment. In the previous versions, if you use a Kubernetes cluster with cgroup v2, you might face an OOM-killed issue.
Bug fixes
ScalarDB Cluster
- Upgraded
grpc_health_probe
to fix security issues. CVE-2024-24790, CVE-2023-45283, and CVE-2023-45288
ScalarDB SQL
- Fixed a bug where incorrect results are returned when executing SELECT queries with the same column names.
- Upgraded
grpc_health_probe
to fix security issues. CVE-2024-24790, CVE-2023-45283, and CVE-2023-45288
v3.10.5
Summary
This release includes several improvements, bug fixes, and vulnerability fixes.
Community edition
Improvements
- Refactored scan with filtering. (#1715)
- Changed the hard-coded password for the Oracle user to a more secure one in the JDBC adapter. (#1765)
- Update base image of container image. This update fixes an OOM issue on a Kubernetes with cgroup v2 environment. In the previous versions, if you use a Kubernetes cluster with cgroup v2, you might face an OOM-killed issue. (#1826)
Bug fixes
- Fixed a bug where
NullPointerException
occurs during theEXTRA_READ
validation when scanning records in a transaction, but some of them are deleted by other transactions. (#1624) - Fixed a bug where users could see inconsistent results when scanning records by an index key after putting the related records in Consensus Commit transactions. (#1727)
- Upgraded
grpc_health_probe
to fix security issues. CVE-2024-24790, CVE-2023-45283, and CVE-2023-45288 (#1980) - Fixed snapshot management issues. (#1976)
Enterprise edition
Improvements
ScalarDB Cluster
- Update base image of container image. This update fixes an OOM issue on a Kubernetes with cgroup v2 environment. In the previous versions, if you use a Kubernetes cluster with cgroup v2, you might face an OOM-killed issue.
ScalarDB GraphQL
- Update base image of container image. This update fixes an OOM issue on a Kubernetes with cgroup v2 environment. In the previous versions, if you use a Kubernetes cluster with cgroup v2, you might face an OOM-killed issue.
ScalarDB SQL
- Update base image of container image. This update fixes an OOM issue on a Kubernetes with cgroup v2 environment. In the previous versions, if you use a Kubernetes cluster with cgroup v2, you might face an OOM-killed issue.
Bug fixes
ScalarDB Cluster
- Upgraded
grpc_health_probe
to fix security issues. CVE-2024-24790, CVE-2023-45283, and CVE-2023-45288
ScalarDB SQL
- Upgraded
grpc_health_probe
to fix security issues. CVE-2024-24790, CVE-2023-45283, and CVE-2023-45288
v3.9.6
Summary
This release includes several improvements, bug fixes, and vulnerability fixes.
Community edition
Improvements
- Changed the hard-coded password for the Oracle user to a more secure one in the JDBC adapter. (#1765)
- Update base image of container image. This update fixes an OOM issue on a Kubernetes with cgroup v2 environment. In the previous versions, if you use a Kubernetes cluster with cgroup v2, you might face an OOM-killed issue. (#1826)
Bug fixes
- Fixed a bug where
NullPointerException
occurs during theEXTRA_READ
validation when scanning records in a transaction, but some of them are deleted by other transactions. (#1624) - Upgraded
grpc_health_probe
to fix security issues. CVE-2024-24790, CVE-2023-45283, and CVE-2023-45288 (#1980)
Enterprise edition
Improvements
ScalarDB Cluster
- Update base image of container image. This update fixes an OOM issue on a Kubernetes with cgroup v2 environment. In the previous versions, if you use a Kubernetes cluster with cgroup v2, you might face an OOM-killed issue.
ScalarDB GraphQL
- Update base image of container image. This update fixes an OOM issue on a Kubernetes with cgroup v2 environment. In the previous versions, if you use a Kubernetes cluster with cgroup v2, you might face an OOM-killed issue.
ScalarDB SQL
- Update base image of container image. This update fixes an OOM issue on a Kubernetes with cgroup v2 environment. In the previous versions, if you use a Kubernetes cluster with cgroup v2, you might face an OOM-killed issue.
Bug fixes
ScalarDB Cluster
- Upgraded
grpc_health_probe
to fix security issues. CVE-2024-24790, CVE-2023-45283, and CVE-2023-45288
ScalarDB SQL
- Upgraded
grpc_health_probe
to fix security issues. CVE-2024-24790, CVE-2023-45283, and CVE-2023-45288
v3.8.7
Summary
This release includes several improvements, bug fixes, and vulnerability fixes.
Improvements
- Changed the hard-coded password for the Oracle user to a more secure one in the JDBC adapter. (#1765)
- Update base image of container image. This update fixes an OOM issue on a Kubernetes with cgroup v2 environment. In the previous versions, if you use a Kubernetes cluster with cgroup v2, you might face an OOM-killed issue. (#1826)
Bug fixes
- Fixed a bug where
NullPointerException
occurs during theEXTRA_READ
validation when scanning records in a transaction, but some of them are deleted by other transactions. (#1624) - Upgraded
grpc_health_probe
to fix security issues. CVE-2024-24790, CVE-2023-45283, and CVE-2023-45288 (#1980)
v3.12.2
Summary
This release includes several improvements including error message improvements and a vulnerability fix.
Community edition
Improvements
- Added error codes to the error messages of Schema Loader. (#1564)
Enterprise edition
Improvements
ScalarDB Cluster
- Added error codes to the error messages of the Auth module.
- Added error codes to the error messages.
- Added TLS support for the Prometheus exporter. With this change, when enabling TLS (setting
scalar.db.cluster.tls.enabled
totrue
) in ScalarDB cluster nodes, the Prometheus exporter also starts with TLS (HTTPS).
ScalarDB GraphQL
- Added error codes to the error messages.
ScalarDB SQL
- Added error codes to the error messages.
Bug fixes
ScalarDB Cluster
- Upgraded the Kubernetes Client Java lib to fix security issues: CVE-2024-25710 and CVE-2024-26308.
v3.11.2
Summary
This release includes several improvements and a vulnerability fix.
Enterprise edition
Bug fixes
ScalarDB Cluster
- Upgraded the Kubernetes Client Java lib to fix security issues: CVE-2024-25710 and CVE-2024-26308.
v3.10.4
Summary
This release includes several improvements and a vulnerability fix.
Enterprise edition
Bug fixes
ScalarDB Cluster
- Upgraded the Kubernetes Client Java lib to fix security issues: CVE-2024-25710 and CVE-2024-26308.
v3.9.5
Summary
This release includes several improvements and a vulnerability fix.
Enterprise edition
Bug fixes
ScalarDB Cluster
- Upgraded the Kubernetes Client Java lib to fix security issues: CVE-2024-25710 and CVE-2024-26308.