GridGain Developers Hub

Monitoring and CLI Changes

This section covers changes to the monitoring stack (logs, metrics, system views, events) and to the command-line tools (control.sh, ignite.sh, sqlline, ignite-cdc.sh).

You can skip it if you have no monitoring, metrics, or log-parsing tooling and no operational scripts or automation to carry over.

Monitoring

  • Logs: the shipped Log4j config format differs, IGNITE_QUIET defaults to false (more verbose), and the log file-name pattern changed. Update log-parsing tooling.

  • Metrics: Ignite.metrics() and the new MetricRegistry API are unavailable. Use JMX or ReadOnlyMetricRegistry.

  • System views: several Ignite-only views (node attributes/metrics, baseline, SQL plan history, snapshot) are unavailable. Query the SYS schema or use JMX.

  • Events: event code 149 means different things in the two products, and several Ignite-only events (snapshot, consistency, SQL query execution) are unavailable. Update event listeners that filter by code.

CLI Changes

The command-line tools are mostly compatible, but several commands and mechanisms differ. See Control Script for the full control.sh reference.

control.sh

These control.sh commands, subcommands, and arguments changed:

Category Command or item Change and what to do

Removed command

--snapshot

Use snapshot-utility.sh.

--cdc

Not supported. Use Data Center Replication instead.

--consistency

Use --cache partition_reconciliation.

--system-view

Use the SYS schema or JMX.

--performance-statistics

Not supported. Use JMX metrics or an external APM.

--cache subcommand

create

Removed. Define the cache in node configuration (XML), or create a table with SQL CREATE TABLE DDL.

scan

Removed. Use a SQL SELECT, or a ScanQuery through the cache API.

metrics

Removed. Enable and read cache metrics over JMX (CacheMetricsMXBean) instead.

idle_verify_dump

Renamed. Call idle_verify --dump instead.

Argument

--ssl-factory

Not available.

--enable-experimental

Set the IGNITE_ENABLE_EXPERIMENTAL_COMMAND system property instead.

Changed output

--state

Prints Cluster is active, Cluster is inactive, or Cluster is active (read-only) instead of Ignite’s Cluster state: ACTIVE format. Update scripts that parse the output.

ignite.sh

Apache Ignite signals a successful restart by writing a file (RESTART_SUCCESS_FILE, set with -DIGNITE_SUCCESS_FILE); GridGain signals it with process exit code 250. Update custom restart wrappers to test for exit code 250 instead of the file. The JVM --add-opens set and the shipped Log4j2 configuration also differ.

sqlline

  • Pass custom JVM options in SQL_JVM_OPTS. Apache Ignite uses JVM_OPTS.

  • The command-history file location changed: Apache Ignite sets --historyFile to ~/.sqlline/ignite_history; GridGain leaves the sqlline default.

ignite-cdc.sh

Not available. GridGain has no CDC at all. For cross-cluster replication, use Data Center Replication; for other consumers, see the CDC entry in Features to Replace.

© 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 12, 2026