GridGain 9.1.23 Release Notes

Overview

GridGain 9.1.23 adds disk-usage write protection, partition awareness in the C++ client, server-side filters for continuous queries, and cancellation tokens for compute jobs. It also adds memory quota block size in JDBC and .NET, local snapshot listing in the CLI, .NET 10 support in the compute executor, and new metrics for storage, thread pools, and the Raft log.

Major Changes

Log File Rollover Preserves Files Across Restarts

The default java.util.logging configuration now sets the FileHandler.append property to true. As a result, a node restart no longer triggers a log generation rollover. The pre-restart log stays in the current generation. Rollover now happens only when a file reaches the size limit (100 MB for the server log, 10 MB for the CLI log by default).

If you parse logs by generation number, update your scripts to account for changed behavior.

New Features

Storage Drive Usage Limits

To protect a node from running out of disk space, GridGain 9.1.23 introduces hard drive-usage limits on both the Raft log and the partitions data directories. When usage on the local drive exceeds the configured limit - or when a quorum-breaking subset of a partition’s peers report the same - incoming client writes for affected partitions are rejected with StorageFullException exception until usage drops back below the limit.

By default, the limits are set to Long.MAX_VALUE, effectively disabling them. You can reduce the limits in the node configuration:

ignite.raft.logStorage {
    hardPartitionsLogDriveUsageLimitBytes = 9223372036854775807
    hardPartitionsDataDriveUsageLimitBytes = 9223372036854775807
}

See hard disk usage limits for details.

Raft Log Storage Size Limits

GridGain 9.1.23 introduces configurable soft limits on the size of Raft log storage. When the size of the persistent partitions' log files (or the volatile partitions' spillout files) exceeds its soft limit, the node triggers Raft snapshots on the affected nodes to truncate logs and reclaim disk space. These limits keep log growth bounded without rejecting client work.

You can configure soft limits in node configuration:

ignite.raft.logStorage {
    softPartitionsLogSizeLimitBytes = 21474836470
    softPartitionsLogSpilloutSizeLimitBytes = 21474836470
}

See soft disk usage limits for details.

C++ Client: Partition Awareness

The C++ client now supports partition-aware routing for single-record operations. When enabled, the client computes the target partition for each key locally and routes the request directly to the primary node, avoiding an extra network hop for misrouted requests.

The example below shows how you can configure your requests to use partition awareness:

auto preferred = ignite::detail::make_typed_preferred_node_fn(
    table_impl, key_tuple);

auto value = table.get(transaction, key_tuple, preferred).get();

For more information, see C++ client documentation.

Partition-Operation Backpressure

A new cluster configuration option limits the percentage of node heap memory that the partition request-processing queue is allowed to occupy. When the limit is exceeded, the node stops admitting new partition operations until in-flight work drains, preventing a single hot node from accepting more requests than it can hold in memory.

ignite.replication {
    partitionOperationHeapUsagePercent = 20   // default: 20 (percent of max JVM heap)
}

Server-Side Remote Filter for Continuous Queries

Continuous queries can now filter events on the server side, reducing network traffic when subscribers are only interested in a subset of changes. The filter is evaluated before events leave the source node.

Java. Set a Criteria-based filter on ContinuousQueryOptions and reference event values through three virtual table names - CUR, OLD, and EVENT:

var options = ContinuousQueryOptions.builder()
    .remoteFilter(Criteria.and(
        Criteria.columnValue("EVENT", "TYPE_ID",
            Criteria.equalTo(TableRowEventType.UPDATED.id())),
        Criteria.columnValue("CUR", "STATUS", Criteria.equalTo("active")),
        Criteria.columnValue("OLD", "PRICE", Criteria.greaterThan(300))))
    .build();

view.queryContinuously(subscriber, options);

In addition to user columns, the EVENT pseudo-table exposes PARTITION_ID (BIGINT), TYPE_ID (INT: 0=CREATED, 1=UPDATED, 2=REMOVED, 3=ARCHIVED), and COMMIT_TIMESTAMP (BIGINT Unix-time milliseconds).

.NET. Two new QueryContinuouslyAsync overloads expose the same filter - one taking a SQL string with positional parameters, and one taking a typed LINQ expression that is translated to SQL on the client:

// SQL form
await foreach (var batch in view.QueryContinuouslyAsync(
    "CUR.PRICE > ? AND EVENT.TYPE_ID = ?",
    [300, (int)TableRowEventType.Updated]))
{
    // process batch
}

// LINQ form
await foreach (var batch in view.QueryContinuouslyAsync(
    x => x.Entry.Value.Price > 300))
{
    // process batch
}

The LINQ overload is not available in AOT-compiled .NET builds; use the SQL form there.

See remote filter in the continuous queries documentation.

Support for Listing Local Snapshot Paths

The snapshot list CLI command can now list the locally stored snapshots. Each node reports the snapshots it holds locally and the results are aggregated for display.

See CLI tool documentation.

Compute Job Cancellation Token

The compute API now exposes the cancellation token to running jobs. Jobs can now obtain the same CancellationToken that triggers their cancellation and forward it to APIs that accept one. For example, you can propogate cancellation to SQL statements:

public class MyJob implements ComputeJob<MyArg, MyResult> {
    @Override
    public CompletableFuture<MyResult> executeAsync(JobExecutionContext ctx, MyArg arg) {
        return ctx.ignite().sql().executeAsync(
                null,
                ctx.cancellationToken(),  // propagates job cancel to SQL
                "SELECT ...");
    }
}

See propagating cancellation from a job for more information.

Memory Quota Block Size for JDBC and .NET

The memoryQuotaBlockSize option introduced for the Java client in 9.1.22 is now available in JDBC and .NET clients. Setting a larger block size reduces synchronization overhead with the node-level memory tracker for memory-intensive queries.

JDBC

The option is exposed on IgniteJdbcStatement:

IgniteJdbcStatement stmt = conn.createStatement().unwrap(IgniteJdbcStatement.class);
stmt.setMemoryQuotaBlockSize(1_048_576L);  // 1 MB blocks

.NET.

A nullable MemoryQuotaBlockSize init-only property is exposed on SqlStatement:

var stmt = new SqlStatement("SELECT * FROM large_table")
{
    MemoryQuotaBlockSize = 1_048_576L  // 1 MB blocks
};

In both clients, the default is null (use the node-level setting). Negative values are rejected. See memory quota block size for more information.

New Metrics

This release adds three new metrics groups for storage, thread-pool, and Raft-log observability. The following metrics were added:

Storage consistency (storage.aipersist.consistency):
- RunConsistentlyDuration - the time spent in runConsistently closures, in nanoseconds.
- RunConsistentlyStarted - the total number of runConsistently invocations started.
- RunConsistentlyActiveCount - current number of active runConsistently calls.
Striped thread-pool aggregated metrics:
- ConcurrencyLevel - number of stripes in the executor.
- ActiveCount - approximate total number of threads currently executing tasks across all stripes.
- IdleCount - approximate number of idle threads across all stripes.
- QueueSize - current total size of the execution queue across all stripes.
Raft log storage:
- PartitionsLogStorageSpilloutSize - the number of bytes occupied on disk by the spillout of volatile-zone partition-group logs.

.NET 10 Support in Compute Executor

With this release, the .NET compute job executor supports .NET 10, so a server node can host .NET compute jobs on either runtime. See .NET compute jobs for more information.

Improvements and Fixed Issues

Issue ID Category Description

Issue ID	Category	Description
IGN-30807	General	Fixed a rare network connection issue.
IGN-30800	Cluster SQL Engine	Added table-scope for SQL hints. Added support for hints' overriding.
IGN-30795	General	Reduced the amount of logs generated when a transaction is handled while a node is stopping.
IGN-30791	General	Fixed connection errors when a cluster node restarts.
IGN-30787	Cluster Metrics & Monitoring	Removed PageAcquireTime metric.
IGN-30767	Cluster SQL Engine	Fixed an issue preventing partition pruning to apply on predicates over dynamic parameters of temporal types.
IGN-30721	Platforms & Clients	Java client: fixed race condition and potential data corruption when decoding large responses.
IGN-30704	General	Fixed non-legitimate "Lease update invocation failed because of outdated lease data" exceptions on leases prolongations causing leases interruptions.
IGN-30657	Distributed Computing	Future returned from a compute job is not automatically canceled to allow normal completion even if a cancellation was requested.
IGN-30626	Platforms & Clients	Client connector: fixed correctness and performance of stack trace reporting in sendServerExceptionStackTraceToClient=true mode.
IGN-30622	Platforms & Clients	Fixed negative TransactionsActive metric for direct transactions.
IGN-30619	Platforms & Clients	Java and .NET clients: never retry authentication errors.
IGN-30446	Cluster Storage Engine	Added new storage consistency metrics.
IGN-30425	Cluster SQL Engine	Fixed client observable timestamp skew in multistatement query cases.
IGN-30328	Platforms & Clients	Java client: reduced verbosity of error stack traces.
IGN-30282	Platforms & Clients	Java client: fixed piggyback transaction rollback.
IGN-30197	Distributed Computing	Added a callback for cancellation API.
IGN-30196	Distributed Computing	Cancellation token is passed to the job execution context to allow streamlined usage of other APIs that take it as an argument.
IGN-28683	Platforms & Clients	Python DB API Driver: Implemented heartbeats.
IGN-28412	General	REST API tags for disaster recovery and system disaster recovery were fixed.
IGN-28155	General	Improved latency for transactions which span multiple partitions
IGN-24461	Platforms & Clients	C++ client: Added support for partition awareness.
GG-48879	General	Fixed a race in low watermark logic.
GG-48796	General	Fixed a possible garbage-collected timestamp in outdated read-only transactions and improved related logging.
GG-48759	Platforms & Clients	Added .NET 10 support to the compute executor.
GG-48744	Platforms & Clients	Fixed .NET runtime version in -dotnet10 Docker images.
GG-48657	Cluster SQL Engine	Fixed an issue causing duplicate rows to be returned when a HASH index is used in query processing.
GG-48612	Cluster SQL Engine	Fixed an issue executing a correlated subquery with the SORT operator.
GG-48536	Cluster SQL Engine	Fixed an issue where queries utilizing the sort aggregate could fail with "AssertionError: Input not sorted".
GG-48449	Cluster SQL Engine	Fixes issue with virtual columns in correlated subqueries.
GG-48444	CLI Tool	Fixed cluster unit deploy ignoring authentication credentials from the CLI secret config file.
GG-48421	Cluster SQL Engine	Fixed an issue with executing a correlated subquery with the LIMIT operator.
GG-48378	Cluster Security	Fixed error that happened when calling Tables API from another thread in the compute job.
GG-48351	Configuration	Added hardLogSizeLimitBytes configuration.
GG-48349	Cluster Metrics & Monitoring	Added raft.log.storage.PartitionsLogStorageSpilloutSize metric.
GG-48325	Distributed Computing	Fixed IgniteCluster.nodesAsync() to return the logical topology (matching nodes()) instead of physical topology, so nodes that have left the cluster or not yet fully joined are no longer included.
GG-48294	Platforms & Clients	C++ Client: Fixed a bug that was blocking GridGain from setting column value to nullptr explicitly.
GG-48284	Builds and Deliveries	Update Spring Boot dependency from 3.5.13 to 3.5.14.
GG-48275	Cluster Data Snapshots and Recovery	Fixed encrypted snapshot buffers allocation.
GG-48234	Distributed Computing	.NET: Added IJobExecutionContext.LoggerFactory and IDataStreamerReceiverContext.LoggerFactory to write to the node log from .NET compute jobs and receivers.
GG-48232	Platforms & Clients	Fixed race condition and improved performance of client message decoding on large payloads.
GG-48226	General	Fixed possible data loss when a disk with Raft log is out of free space.
GG-48215	Platforms & Clients	C++ Client: Added examples.
GG-48200	Cluster Metrics & Monitoring	Added aggregated metrics for a striped thread pool.
GG-48193	Platforms & Clients	Fixed an issue caused ReadOnly flag on JDBC Connection to be ignored.
GG-48179	Cluster Data Replication	Fixed DCR error handling.
GG-48175	Platforms & Clients	Thin clients: fixed potential update visibility issues on implicit tx retry.
GG-48167	Platforms & Clients	Added reporting of multi-statement execution errors when the user closes a Statement without iterating through all results.
GG-48165	Cluster Data Replication	DCR: deletes made on the source while replication was offline are now applied on the target after restart.
GG-48145	Cluster Storage Engine	Fixed a rare partition corruption that occurred when primary replica is lagging or being restarted under load.
GG-48137	General	Fixed an error when a node could not join the cluster after cluster rename.
GG-48109	Configuration	Added new configuriton option partitionOperationHeapUsagePercent that limits the percentage of node heap memory allocated for the user request processing queue, preventing a node from accepting too many operations and exhausting available memory.
GG-48091	Builds and Deliveries	Log files now roll over only when the maximum file size is exceeded, and existing log files are preserved across restarts.
GG-48031	Builds and Deliveries	Added Java 26 support to Docker images.
GG-47978	Cluster SQL Engine	Improved warning for unsupported COPY INTO direction.
GG-47976	Cluster SQL Engine	Fixed COPY INTO command allowing to bypass user permissions.
GG-47951	Cluster Data Replication	Java client: fixed exception message on auth failure.
GG-47949	Cluster Data Replication	DCR replications stuck in the replicating state are now recovered when worker node is restarted.
GG-47940	General	Fixed a race where rolled-back transactions could block a partition by leaving behind a shared lock (S-lock).
GG-47934	General	Implicit transactions can now be retried in case of unexpected retriable intercluster network issues
GG-47916	CLI Tool	Improved UX when connecting to a node during the initialization process.
GG-47905	Platforms & Clients	Java client: Fixed initial connection fail if retry policy is null and the first address is unavailable.
GG-47870	Cluster Data Replication	DCR is now retried correctly if source cluster is not available.
GG-47866	Cluster Data Snapshots and Recovery	Fixed data inconsistency when restoring snapshot on a zone with multiple replicas.
GG-47834	Configuration	Old configurations are now properly migrated.
GG-47820	Cluster Storage Engine	Metric PageAcquireTime is deprecated and will always have 0 value until removed. To be removed in 9.2.
GG-47810	Configuration	Some configuration values from old versions are now properly migrated.
GG-47809	Cluster SQL Engine	Added support for non-default schema in COPY FROM command.
GG-47807	Cluster Security	RBAC: corrected username display to use the actual username instead of the display name.
GG-47805	Platforms & Clients	Java client: fixed connection leak when client initialization fails.
GG-47787	General	REST endpoints no longer return 409 while cluster is initializing.
GG-47747	Cluster Continuous Queries	Added remote filter for continuous queries.
GG-47720	Cluster Continuous Queries	Continuous Query: optimized long polling logic.
GG-47668	Migration Tools	Ignore attribute is no longer skipped on key fields in cache mapping.
GG-47591	Platforms & Clients	.NET: Added Statement.MemoryQuotaBlockSize.
GG-47580	Platforms & Clients	Memory Quota Block Size option is now available in JDBC Statements.
GG-47578	Cluster Storage Engine	Fixed rare index build failures when many indexes are created at the same time.
GG-47557	Cluster Security	Optimized number of heap allocations in RBAC.
GG-47544	CLI Tool	Improved user list command output.
GG-47543	Cluster Security	Added support for ALTER USER SQL command.
GG-47541	Cluster Storage Engine	GridGain now pauses client writes when configured storage size limits are exceeded.
GG-47539	General	Added soft limits for log storage that trigger log storage compaction.
GG-47455	Cluster Security	You can now specify multiple user attributes in LDAP.
GG-47384	Cluster Data Snapshots and Recovery	Fixed error when expiration runs during snapshot restoration.
GG-47262	Cluster Continuous Queries	.NET: Continuous Query: added overloads with `remoteFilter` - SQL expression and LINQ expression - to filter events on the server side.
GG-47141	Platforms & Clients	Added JDBC driver property that permits running DDL statements in a session with active explicit transaction.
GG-47093	Platforms & Clients	Implicit transactions with SQL will be retried in case primary replica expiration during the execution.
GG-46575	CLI Tool	Added a new CLI command to list local snapshots.
GG-46396	Builds and Deliveries	Added new docker images for GG9 based on amazoncorretto:17-al2023 and rockylinux:9.
GG-44871	Cluster Continuous Queries	.NET: Added ContinuousQueryOptions.SkipOldEntries.
GG-44065	Cluster Continuous Queries	Continuous Query: added remoteFilter option to filter events on the server side.
GG-38308	Cluster Data Snapshots and Recovery	Missing schemas are now automatically created when restoring a snapshot

IGN-30807

General

Fixed a rare network connection issue.

IGN-30800

Cluster SQL Engine

Added table-scope for SQL hints. Added support for hints' overriding.

IGN-30795

General

Reduced the amount of logs generated when a transaction is handled while a node is stopping.

IGN-30791

General

Fixed connection errors when a cluster node restarts.

IGN-30787

Cluster Metrics & Monitoring

Removed PageAcquireTime metric.

IGN-30767

Cluster SQL Engine

Fixed an issue preventing partition pruning to apply on predicates over dynamic parameters of temporal types.

IGN-30721

Platforms & Clients

Java client: fixed race condition and potential data corruption when decoding large responses.

IGN-30704

General

Fixed non-legitimate "Lease update invocation failed because of outdated lease data" exceptions on leases prolongations causing leases interruptions.

IGN-30657

Distributed Computing

Future returned from a compute job is not automatically canceled to allow normal completion even if a cancellation was requested.

IGN-30626

Platforms & Clients

Client connector: fixed correctness and performance of stack trace reporting in sendServerExceptionStackTraceToClient=true mode.

IGN-30622

Platforms & Clients

Fixed negative TransactionsActive metric for direct transactions.

IGN-30619

Platforms & Clients

Java and .NET clients: never retry authentication errors.

IGN-30446

Cluster Storage Engine

Added new storage consistency metrics.

IGN-30425

Cluster SQL Engine

Fixed client observable timestamp skew in multistatement query cases.

IGN-30328

Platforms & Clients

Java client: reduced verbosity of error stack traces.

IGN-30282

Platforms & Clients

Java client: fixed piggyback transaction rollback.

IGN-30197

Distributed Computing

Added a callback for cancellation API.

IGN-30196

Distributed Computing

Cancellation token is passed to the job execution context to allow streamlined usage of other APIs that take it as an argument.

IGN-28683

Platforms & Clients

Python DB API Driver: Implemented heartbeats.

IGN-28412

General

REST API tags for disaster recovery and system disaster recovery were fixed.

IGN-28155

General

Improved latency for transactions which span multiple partitions

IGN-24461

Platforms & Clients

C++ client: Added support for partition awareness.

GG-48879

General

Fixed a race in low watermark logic.

GG-48796

General

Fixed a possible garbage-collected timestamp in outdated read-only transactions and improved related logging.

GG-48759

Platforms & Clients

Added .NET 10 support to the compute executor.

GG-48744

Platforms & Clients

Fixed .NET runtime version in -dotnet10 Docker images.

GG-48657

Cluster SQL Engine

Fixed an issue causing duplicate rows to be returned when a HASH index is used in query processing.

GG-48612

Cluster SQL Engine

Fixed an issue executing a correlated subquery with the SORT operator.

GG-48536

Cluster SQL Engine

Fixed an issue where queries utilizing the sort aggregate could fail with "AssertionError: Input not sorted".

GG-48449

Cluster SQL Engine

Fixes issue with virtual columns in correlated subqueries.

GG-48444

CLI Tool

Fixed cluster unit deploy ignoring authentication credentials from the CLI secret config file.

GG-48421

Cluster SQL Engine

Fixed an issue with executing a correlated subquery with the LIMIT operator.

GG-48378

Cluster Security

Fixed error that happened when calling Tables API from another thread in the compute job.

GG-48351

Configuration

Added hardLogSizeLimitBytes configuration.

GG-48349

Cluster Metrics & Monitoring

Added raft.log.storage.PartitionsLogStorageSpilloutSize metric.

GG-48325

Distributed Computing

Fixed IgniteCluster.nodesAsync() to return the logical topology (matching nodes()) instead of physical topology, so nodes that have left the cluster or not yet fully joined are no longer included.

GG-48294

Platforms & Clients

C++ Client: Fixed a bug that was blocking GridGain from setting column value to nullptr explicitly.

GG-48284

Builds and Deliveries

Update Spring Boot dependency from 3.5.13 to 3.5.14.

GG-48275

Cluster Data Snapshots and Recovery

Fixed encrypted snapshot buffers allocation.

GG-48234

Distributed Computing

.NET: Added IJobExecutionContext.LoggerFactory and IDataStreamerReceiverContext.LoggerFactory to write to the node log from .NET compute jobs and receivers.

GG-48232

Platforms & Clients

Fixed race condition and improved performance of client message decoding on large payloads.

GG-48226

General

Fixed possible data loss when a disk with Raft log is out of free space.

GG-48215

Platforms & Clients

C++ Client: Added examples.

GG-48200

Cluster Metrics & Monitoring

Added aggregated metrics for a striped thread pool.

GG-48193

Platforms & Clients

Fixed an issue caused ReadOnly flag on JDBC Connection to be ignored.

GG-48179

Cluster Data Replication

Fixed DCR error handling.

GG-48175

Platforms & Clients

Thin clients: fixed potential update visibility issues on implicit tx retry.

GG-48167

Platforms & Clients

Added reporting of multi-statement execution errors when the user closes a Statement without iterating through all results.

GG-48165

Cluster Data Replication

DCR: deletes made on the source while replication was offline are now applied on the target after restart.

GG-48145

Cluster Storage Engine

Fixed a rare partition corruption that occurred when primary replica is lagging or being restarted under load.

GG-48137

General

Fixed an error when a node could not join the cluster after cluster rename.

GG-48109

Configuration

Added new configuriton option partitionOperationHeapUsagePercent that limits the percentage of node heap memory allocated for the user request processing queue, preventing a node from accepting too many operations and exhausting available memory.

GG-48091

Builds and Deliveries

Log files now roll over only when the maximum file size is exceeded, and existing log files are preserved across restarts.

GG-48031

Builds and Deliveries

Added Java 26 support to Docker images.

GG-47978

Cluster SQL Engine

Improved warning for unsupported COPY INTO direction.

GG-47976

Cluster SQL Engine

Fixed COPY INTO command allowing to bypass user permissions.

GG-47951

Cluster Data Replication

Java client: fixed exception message on auth failure.

GG-47949

Cluster Data Replication

DCR replications stuck in the replicating state are now recovered when worker node is restarted.

GG-47940

General

Fixed a race where rolled-back transactions could block a partition by leaving behind a shared lock (S-lock).

GG-47934

General

Implicit transactions can now be retried in case of unexpected retriable intercluster network issues

GG-47916

CLI Tool

Improved UX when connecting to a node during the initialization process.

GG-47905

Platforms & Clients

Java client: Fixed initial connection fail if retry policy is null and the first address is unavailable.

GG-47870

Cluster Data Replication

DCR is now retried correctly if source cluster is not available.

GG-47866

Cluster Data Snapshots and Recovery

Fixed data inconsistency when restoring snapshot on a zone with multiple replicas.

GG-47834

Configuration

Old configurations are now properly migrated.

GG-47820

Cluster Storage Engine

Metric PageAcquireTime is deprecated and will always have 0 value until removed. To be removed in 9.2.

GG-47810

Configuration

Some configuration values from old versions are now properly migrated.

GG-47809

Cluster SQL Engine

Added support for non-default schema in COPY FROM command.

GG-47807

Cluster Security

RBAC: corrected username display to use the actual username instead of the display name.

GG-47805

Platforms & Clients

Java client: fixed connection leak when client initialization fails.

GG-47787

General

REST endpoints no longer return 409 while cluster is initializing.

GG-47747

Cluster Continuous Queries

Added remote filter for continuous queries.

GG-47720

Cluster Continuous Queries

Continuous Query: optimized long polling logic.

GG-47668

Migration Tools

Ignore attribute is no longer skipped on key fields in cache mapping.

GG-47591

Platforms & Clients

.NET: Added Statement.MemoryQuotaBlockSize.

GG-47580

Platforms & Clients

Memory Quota Block Size option is now available in JDBC Statements.

GG-47578

Cluster Storage Engine

Fixed rare index build failures when many indexes are created at the same time.

GG-47557

Cluster Security

Optimized number of heap allocations in RBAC.

GG-47544

CLI Tool

Improved user list command output.

GG-47543

Cluster Security

Added support for ALTER USER SQL command.

GG-47541

Cluster Storage Engine

GridGain now pauses client writes when configured storage size limits are exceeded.

GG-47539

General

Added soft limits for log storage that trigger log storage compaction.

GG-47455

Cluster Security

You can now specify multiple user attributes in LDAP.

GG-47384

Cluster Data Snapshots and Recovery

Fixed error when expiration runs during snapshot restoration.

GG-47262

Cluster Continuous Queries

.NET: Continuous Query: added overloads with remoteFilter - SQL expression and LINQ expression - to filter events on the server side.

GG-47141

Platforms & Clients

Added JDBC driver property that permits running DDL statements in a session with active explicit transaction.

GG-47093

Platforms & Clients

Implicit transactions with SQL will be retried in case primary replica expiration during the execution.

GG-46575

CLI Tool

Added a new CLI command to list local snapshots.

GG-46396

Builds and Deliveries

Added new docker images for GG9 based on amazoncorretto:17-al2023 and rockylinux:9.

GG-44871

Cluster Continuous Queries

.NET: Added ContinuousQueryOptions.SkipOldEntries.

GG-44065

Cluster Continuous Queries

Continuous Query: added remoteFilter option to filter events on the server side.

GG-38308

Cluster Data Snapshots and Recovery

Missing schemas are now automatically created when restoring a snapshot

Upgrade Information

You can upgrade to current GridGain version from previous releases. Below is a list of versions that are compatible with the current version. Compatibility with other versions is not guaranteed. If you are on a version that is not listed, contact GridGain for information on upgrade options.

9.1.8, 9.1.9, 9.1.10, 9.1.11, 9.1.12, 9.1.13, 9.1.14, 9.1.15, 9.1.16, 9.1.17, 9.1.18, 9.1.19, 9.1.20, 9.1.21, 9.1.22

When updating from older versions, we recommend updating to version 9.1.8 first, before performing an update to current version.

Known Limitations

Rolling Upgrade Over 9.1.10

When performing a rolling upgrade from GridGain 9.1.9 or earlier, it is necessary to first upgrade to 9.1.10 before upgrading to 9.1.11 or a later version. This is caused by improvements in rolling upgrade procedure that make skipping 9.1.10 with a rolling upgrade impossible.

The upgrade to 9.1.10 must be performed for the whole cluster and committed before the next upgrade can be started.

You do not need to perform this intermediary upgrade when upgrading with downtime, as it only affects the rolling upgrade procedure.

Data Restoration After Data Rebalance

Currently, data rebalance may cause partition distribution to change and cause issues with snapshots and data recovery. In particular:

It is currently not possible to restore a LOCAL snapshot if data rebalance happened after snapshot creation. This will be addressed in one of the upcoming releases.
It is currently not possible to perform point-in-time recovery if data rebalance happened after table creation. This will be addressed in one of the upcoming releases.

SQL Performance in Complex Scenarios

There are known issues with the performance of SQL read-write transactions in complex read-write scenarios. These issues will be addressed in upcoming releases.

We Value Your Feedback

Your comments and suggestions are always welcome. You can reach us here: http://support.gridgain.com/.

© 2026 GridGain Systems, Inc. All Rights Reserved. Privacy Policy | Legal Notices. GridGain® is a registered trademark of GridGain Systems, Inc.
Apache, Apache Ignite, the Apache feather and the Apache Ignite logo are either registered trademarks or trademarks of The Apache Software Foundation.

Last updated on Jun 02, 2026