Individual graphs are defined below.

avg_row_accesses.name

Shows the average of the number of accesses for data for different query operations for all table reads, regardless of the storage engine being using, against the number of queries being executed.

binlog_cache.name

binlog_space_usage.name

cluster_data_node_data_memory_usage.name

cluster_data_node_index_memory_usage.name

cluster_data_node_redo_buffer_usage.name

cluster_data_node_redo_logspace_usage.name

cluster_data_node_undo_buffer_usage.name

cluster_data_node_undo_logspace_usage.name

connections.name

Displays the number of client connections by displaying the number of connected threads, active threads, and cached threads.

connections_aborted.name

Shows the number of client connections that have been aborted, either during connection, or aborted active connections. This information is useful to determine if your server is overloaded (and therefore aborting connections), or if there is a network fault. Aborted connections also indicate those which did not disconnect properly, for example during a timeout.

connections_max.name

Shows the maximum number of connections, and the maximum number of used connections.

CPU Utilization

Displays the CPU usage statistics over time for the machine hosting the MySQL server. The graph shows the aggregate system, user, wait and idle times for your CPU(s). You can use this information to help determine your CPU utilization and whether you are getting the most efficient use of your CPU power.

database_activity.name

Displays the database activity, by showing the individual counts over time of different DML statements per minute. For example, you can monitor the number of SELECT statements executed per minute over time and gain an understanding of your server throughput. This information can be used to monitor the activity level, and provide a useful quick reference for your server operation.

database_transactions.name

Disk IO Usage

Disk Space Usage

hit_ratios.name

Shows the cache hit ratios (as a percentage) for the query cache, MyISAM and InnoDB tables, and the connection and thread cache information. The information provided by the hit ratios can provide indicators of whether your cache is working effectively, or is of a size suitable for the data that you are storing and retrieving. Low hit ratios indicate that your cache are inefficient and may need to be investigated.

innodb_adaptive_hash_memory.name

Shows the size of the InnoDB adaptive hash index. These are used to hold the data from tables in memory providing the size of the indexes is small enough to fit in memory. The graph shows the memory usage of the adaptive hash index.

innodb_adaptive_hash_searches.name

Displays the number of InnoDB adaptive hash index searches and those searches not using the adaptive hash index. You can use this to determine if your adaptive hash index is being used correctly, and is of a suitable size of the information and tables that are being accessed.

innodb_buffer_pool.name

Displays the size and page usage of the InnoDB buffer pool, providing information on the cache usage and performance. The graph shows the buffer pool size, active pages, and the number of modified pages.

innodb_compression_time.name

Shows the time spent by InnoDB compressing and uncompressing data in compressed tables.

innodb_hist_length.name

The size of the InnoDB transaction history. A high number may indicate a large number of open (uncommitted) transactions.

innodb_os_file_access.name

Displays the InnoDB I/O counts showing the number of fsync, read, and write operations used to support InnoDB tables.

innodb_row_details.name

Shows the row counts per minute for individual SQL operations (READ, INSERT, UPDATE and DELETE).

innodb_semaphores.name

Displays the InnoDB semaphore status. The graph indicates the amount of time that different InnoDB threads have spent waiting due to the locks used to prevent the same structures being updated at the same time. A large number of threads waiting for the semaphores may be a result of disk I/O or connection problems inside InnoDB. You may want to modify your InnoDB thread concurrency.

innodb_transactions.name

The number of blocked, locked, and active transactions. You can use this to determine the contention on tables for InnoDB transactions.

kbytes_in_out.name

Displays the total Kilobytes per minute of bytes transferred to/from the server by client applications. Spikes in this output may indicate an unusual application operation or connection.

Load Average

The load average of the server hosting the MySQL server. The load average shows the number of processes using or waiting for CPU time in the last 1, 5 and 15 minutes. Load averages showing a load highe than the number of CPU cores may indicate an overloaded server.

lock_wait_ratio.name

Shows the percentage of table locks that have had to wait before an operation can continue. High lock ratios can indicate high contention in your tables, with too many operations having to wait, affecting the performance.

Memory Usage - Agent

Displays the current amount of memory used by the agent, and by the Lua component of the agent.

Memory Usage - OS Resident

Displays the RAM usage on the server for the monitored MySQL instance. You should compare the total and used RAM values to ensure that you are not exceeding your available RAM, which will lead to swapping and performance issues.

Memory Usage - OS Virtual

Displays the use of swap space on the server for the monitored MySQL instance. High swap usage may indicate that your server needs more RAM or that your MySQL configuration needs to be modified, as high levels of swap will have a negative impact on performance.

myisam_key_buffer_activity.name

myisam_key_buffer_usage.name

Displays the efficiency of the MyISAM key index buffer usage. On a system using a lot of MyISAM tables, low usage indicates that your MyISAM key buffer size is too high (reducing the RAM available for other buffering). High usage may indicate that you need to increase your MyISAM key buffer size to help improve performance.

Network - Interface Throughput

opened_tables.name

The number of tables in the open state per minute.

query_cache_efficiency.name

Displays the hits, inserts and queries not cacheable in the query cache. Low cache hits may indicate that your queries are being expired from the cache before they can be used, which may mean they are unsuitable for storing in the query cache.

query_cache_lowmem_prunes.name

Displays the number of queries removed from the cache because the size of the query cache was not large enough to store the queries that can be cached. Try increasing your query cache size.

query_cache_memory.name

Displays the free space and total size of the query cache.

query_cache_queries_in_cache.name

Shows the number of queries stored in the query cache. In normal operation this should be a relatively constant figure. A large number of switches between high and low numbers may indicate that differently sized queries are being inserted into the query cache, and then later removed as a high number of smaller queries are added to the cache. Try increasing the size of your query cache.

row_accesses.name

Displays the aggregated row access statistics per minute. Information is shown both in terms of the full table scans (which are expensive to perform), and index based accesses.

row_writes.name

Shows the number of delete, write (INSERT), and update operations on rows per minute on all tables regardless of storage engine.

slave_seconds_behind.name

Displays the number of seconds behind the master for a given slave in a replication scenario. An increasing value means that your slave is unable to keep up with your master.

sorting.name

Shows the number of different sort operations performed on queries.

table_locks.name

Shows the average number of table locks per second.

temporary_tables.name

Displays the number of memory temporary tables and disk temporary tables. Disk temporary tables are slower to create, populate, and read back. You may want to increase your memory temporary table size or check your queries to determine whether the use of temporary tables can be minimized.

thread_cache.name

Displays the thread cache information, comparing new thread creations against all database connections.