> For the complete documentation index, see [llms.txt](https://mariadb.com/docs/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://mariadb.com/docs/tools/mariadb-enterprise-manager/usage/monitoring/dashboards/mariadb-server.md).

# MariaDB Server

This dashboard provides a unified view of a database topology. It combines topology information, system health, replication or cluster metrics, and query performance in one place. Administrators can use it to monitor availability, troubleshoot issues, and optimize performance.

## Topology Overview

Summarizes the overall topology, showing which servers are active, their versions, and how they are organized.

<figure><img src="/files/KcolZByqbGytoXzn6R75" alt=""><figcaption></figcaption></figure>

* Name — Displays the name of the MariaDB topology currently being monitored.
* Project — Shows the associated project or environment label.
* Primary/Replica — A table with:
  * Instance: Server hostname.
  * Type: Instance role.
  * Seconds behind primary: Replication delay value.
  * Status: Availability of the node.
* Topology Info — Count of nodes grouped by type (e.g., server, MaxScale).
* MariaDB Server Uptime by Instance — Uptime in seconds for each server instance.

## System Metrics

<figure><img src="/files/gU3PcPNFiUjH3ZPVGJZP" alt=""><figcaption></figcaption></figure>

Monitors server resource usage to detect bottlenecks in CPU, memory, network, and storage.

| Feature                     | Description                                                                    |
| --------------------------- | ------------------------------------------------------------------------------ |
| CPU Utilisation             | Line graph of CPU usage percentage per instance.                               |
| Memory Usage                | Percentage of used memory per instance (excluding cache/buffers).              |
| Network Traffic             | Time-series of receive and transmit throughput per instance (bits per second). |
| Filesystems Info            | Table with filesystem type, mount point, capacity, and instance.               |
| Disk Used Space Utilisation | Graph of percentage disk space used per mount point.                           |
| Disk IOPS                   | Reads and writes per second per storage device.                                |

## Replication / Cluster Metrics

<figure><img src="/files/ajL4UTvDfLZsXi1upYen" alt=""><figcaption></figcaption></figure>

Provides insight into replication and cluster-related activity, including binary log usage, commit rates, and delay measurements.

| Metric            | Description                                        |
| ----------------- | -------------------------------------------------- |
| Binlog Size       | Current binary log size per instance.              |
| Binlog Throughput | Bytes written to binary logs per second.           |
| Binlog Commits    | Rate of commit operations recorded in binary logs. |
| Replication Lag   | Replication delay value reported in seconds.       |

**Replication Status Table**

This table provides a consolidated view of the health status of replication across instances.

| Field Name             | Description                                                    |
| ---------------------- | -------------------------------------------------------------- |
| Slave\_connections     | Number of replication I/O connections to the upstream source.  |
| Retried\_transactions  | Total replicated transactions retried due to transient errors. |
| Slave\_IO\_Running     | Status flag indicating if the I/O thread is fetching events.   |
| Slave\_SQL\_Running    | Status flag indicating if the SQL thread is applying events.   |
| Last\_Errno            | Most recent numeric error code for replication issues overall. |
| Last\_IO\_Errno        | Most recent numeric error code reported by the I/O thread.     |
| Last\_SQL\_Errno       | Most recent numeric error code reported by the SQL thread.     |
| Read\_Master\_Log\_Pos | Current read position in the source’s binary log.              |
| Relay\_Log\_Pos        | Last executed position in the local relay log.                 |

## Query Metrics

<figure><img src="/files/MCLsMBZiqsxru2NDd0xj" alt=""><figcaption></figcaption></figure>

Focuses on query execution and workload behavior, highlighting concurrency, throughput, and inefficiencies.

| Metric                  | Description                                   |
| ----------------------- | --------------------------------------------- |
| Current Threads Running | Number of threads actively executing queries. |
| Questions (QPS)         | Queries per second executed on each instance. |
| Slow Queries            | Rate of queries exceeding `long_query_time`.  |
| Created Tmp Disk Tables | On-disk temporary tables created per second.  |

## Connections

<figure><img src="/files/qVW4jAYh5t7medQosvm1" alt=""><figcaption></figcaption></figure>

This section provides visibility into how clients connect to the server and whether connection limits or failures are occurring.

| Metric                   | Description                                                                                             |
| ------------------------ | ------------------------------------------------------------------------------------------------------- |
| Number of Connections    | Current number of active client connections (`Threads_connected`).                                      |
| Connection Utilization   | Share of connections in use compared to the configured maximum (`Threads_connected / max_connections`). |
| % of Aborted Connections | Percentage of connection attempts that failed or were aborted (`aborted_connects / connections`).       |

## Range Metrics

<figure><img src="/files/XMwe8qOQsDBbnjFbKf5Q" alt=""><figcaption></figcaption></figure>

Highlights query access patterns where range operations or scans are used.

| Metric                 | Description                                                                                                     |
| ---------------------- | --------------------------------------------------------------------------------------------------------------- |
| Select Range Scan      | Number of `SELECT` operations performing range scans.                                                           |
| Select Full Range Join | Number of queries that performed a full range join. Indicates potential suboptimal indexing or join conditions. |
| Select Range Check     | Number of `SELECT` operations requiring range checks.                                                           |

## InnoDB Metrics

<figure><img src="/files/6eGsnkapOaVwtLpTbwTv" alt=""><figcaption></figcaption></figure>

Shows activity within the InnoDB storage engine.

| Metric                     | Description                                                                                                       |
| -------------------------- | ----------------------------------------------------------------------------------------------------------------- |
| InnoDB Read/Writes         | Rate of physical read and write operations by InnoDB per second. Reads are disk fetches, writes are disk flushes. |
| InnoDB Buffer Pool Reads   | Logical reads from the buffer pool vs. evicted or read-ahead pages, indicating buffer pool efficiency.            |
| InnoDB Row Lock            | Number of row lock waits in InnoDB, with high values indicating contention or poor indexing.                      |
| InnoDB Checkpoint Age      | Size of uncheckpointed redo log data in bytes, with large sizes signaling risk of long crash recovery times.      |
| InnoDB Log Writes          | Number of write operations to the InnoDB redo log per second, reflecting redo logging activity.                   |
| InnoDB History List Length | Length of the undo log history list, with growth indicating long-running transactions preventing purge.           |
| Deadlocks                  | Number of detected deadlocks, where transactions block each other and require one to be rolled back.              |

## Processlist

Shows information about active sessions and thread states collected from `information_schema.processlist`.

* Processlist Count — Table view showing:
  * Instance: Database node.
  * Client: Client host connected.
  * Value: Number of processes/threads from that client.

<sub>*This page is: Copyright © 2025 MariaDB. All rights reserved.*</sub>

{% @marketo/form formId="4316" %}