MariaDB Xpand Administration Tools

This page outlines some frequently used utilities and tools available within MariaDB Xpand.

This documentation applies to both Xpand topologies.

Connect to Xpand from the Command-line

Connections to MariaDB Xpand can be made from the command-line using MariaDB Client.

When using the Xpand Performance topology, MariaDB Client can connect to the MaxScale nodes and the Xpand nodes.

When using the Xpand Storage Engine topology, MariaDB Client can connect to the MaxScale nodes, the ES nodes, and the Xpand nodes.

For example, to connect to a MaxScale node, ES node, or Xpand node using TCP/IP:

$ mariadb --host=192.0.2.10 \
   --port=3307 \
   --user=xpand_app_user \
   --password

Or to connect to a local ES node or Xpand node using the Unix socket file:

$ mysql --user=xpand_user --password

For additional information, see "Connect to MariaDB Xpand".

Loading Data onto Xpand

Depending on your specific needs, there are numerous options for loading data onto Xpand.

The recommended data load method is to use the clustrix_import utility. The clustrix_import utility can be used if your source database is MariaDB Enterprise Server or if you have a mariadb-dump file available. This is the fastest method for loading data from another system into MariaDB Xpand.

When using both Xpand topologies, clustrix_import must be executed on one the Xpand nodes.

Another data load method is the LOAD DATA INFILE statement. The LOAD DATA INFILE statement can be used to import data from .tsv and .csv files.

When using the Xpand Performance topology, LOAD DATA INFILE statement can be executed using the MaxScale nodes and the Xpand nodes.

When using the Xpand Storage Engine topology, LOAD DATA INFILE statement can be executed using the MaxScale nodes, the ES nodes, and the Xpand nodes.

See Loading Data with Xpand for more information.

Web-based Cluster Administration

Xpand includes an Xpand Administration UI for administration.

Examining Xpand Logs

Xpand's logs allow users and Customer Support to understand the queries and events that occur on an Xpand deployment.

For both Xpand topologies, all the logs created by Xpand are located on the Xpand nodes in the /data/clustrix/log directory by default.

There are a separate set of logs on each Xpand node. The logs used most frequently are query.log and clustrix.log.

The query.log shows information for SQL statements used within Xpand. From the query.log you can identify slow queries as well as query syntax errors. For information see Query Logging.

The clustrix.log is a chronology of events that occur on your deployment. It includes alerts that are also logged to a separate log file - alert.log. Please see Email Alerts to establish the necessary alerts for your deployment.

Using CLX - Cluster-Wide Command Line Utility

Xpand includes a general-purpose command line utility called clx for monitoring and managing a deployment.

For both Xpand topologies, the clx command must be executed on one of the Xpand nodes.

A single clx command is automatically processed in parallel on all Xpand nodes. It is located in the /opt/clustrix/bin/ directory.

The most useful commands are:

Command

Description

clx status

Displays the status, IP, and disk utilization of each node in the deployment.

clx cmd

Runs Linux commands in parallel on all nodes

clx logdump

Consolidates and filters Xpand logs so database issues can be researched and resolved.

clx dbstop

Stops Xpand on all nodes.

clx dbstart

Starts Xpand on all nodes.

clx dbrestart

Restarts Xpand on all nodes. (Use in lieu of dbstop and dbstart where applicable.)

Xpand clx utility also communicates with another critical element of the system called nanny. The nanny is a database deployment process that provides monitoring to ensure that all processes vital to Xpand are running. It is activated at system startup. If any critical Xpand component fails at any point, for any reason, nanny restarts it automatically.

Monitoring Xpand Using statd

Performance statistics are automatically collected by statd in the clustrix_statd database. The resulting statistics can be assessed by querying the tables directly or by viewing them with the Xpand administrative tool.

For a recap of the statistics gathered and some potential ways to analyze them, see Monitoring with statd

Upgrading Xpand

Xpand regularly provides both major and minor releases to continuously improve its product's performance, functionality, scalability, and stability. Each release passes a thorough and rigorous quality assurance process before it is released to our customers.

Expanding/Reducing Capacity

Your application and database may experience fluctuating transaction volumes due to seasonal changes. The underlying design of Xpand enables it to easily and uniquely adapt to those changes. To increase your deployment’s capacity, simply add more Xpand nodes. Conversely, if transaction volume subsequently declines, removing Xpand nodes to decrease costs is equally painless. See Scale-Out and Scale-In for specifics.

Parallel Backup and Restore

Xpand includes a fast parallel backup and restore feature. For backups, each Xpand node sends its data directly to the backup target in parallel, allowing for concurrency and for backup performance to scale with deployment size. Similarly, the restore operation is initiated on one Xpand node, and that Xpand node coordinates with all other Xpand nodes to read from the backup file and restore data throughout the deployment in parallel.

The Xpand database backups contain schema definitions, metadata relative to the backup, compacted row data, and information to ensure data consistency.

See Fast Backup and Restore for more information.