Database monitoring measures

This page lists available measures for database monitoring.

DB2

  • Session count - all databases: Overall number of sessions for all databases.
  • Sessions connected to the monitored database: Number of sessions for monitored databases.
  • Sessions connected to non-monitored databases: Number of sessions for non-monitored databases.
  • Activities - Completed: Average number of successfully completed activities per second. Use this element to determine the throughput of activities in the system. Refer to the DB2 documentation for the following monitor elements:
  • Activities - Rejected: Average number of rejected activities per second. Use this element to determine whether predictive thresholds and work actions that prevent execution are effective, and whether they are too restrictive. Refer to the DB2 documentation for the following monitor elements:
  • Activities - Aborted: Average number of cancelled activities per second, or activities completed with errors. Activities may be cancelled due to cancellation, errors, or reactive thresholds. Refer to the DB2 documentation for the following monitor elements:
  • SQL processing time: Average number of seconds spent on processing SQL per second, including time spent to perform section execution. Refer to the DB2 documentation for the following monitor elements:
  • Commit and rollback processing time: Average time spent on commits and rollbacks per second, including time spent to commit processing and rollback operations on the database server. Refer to the DB2 documentation for the following monitor elements:
  • Administration tasks processing time: Average time spent on administration tasks per second, including time spent to perform run stats, reorg, load processing, online backups, and index builds on database server. Refer to the DB2 documentation for the following monitor elements:
  • User routines processing time, Average number time spent on executing routines outside of known DB2 per second. Refer to the DB2 documentation for the following monitor elements:
  • Compilation time: Average time spent on compilation per second, including time spent performing explicit and implicit compiles on the database server. Refer to the DB2 documentation for the following monitor elements:
    • TOTAL_COMPILE_PROC_TIME
    • TOTAL_IMPLICIT_COMPILE_PROC_TIME
      Explicit compiles are compilations directly initiated by a user request such as a bind, rebind, prepare or execute immediate. Implicit compiles are compilations not directly requested by the user. For example, an implicit compilation may occur when executing a statement that was bound using the VALIDATE RUN option if the statement needs to be compiled at execution time.
  • Connection management time: Average time spent on connection management per second, including time spent processing a connection or switch user request. Refer to the DB2 documentation for the following monitor elements:
  • CPU time: Average amount of CPU time used within DB2 per second. It includes both user and system CPU time. Refer to the DB2 documentation for the following monitor elements:
  • Read time: Average time per second spent to:
    • Read in data and index pages from the table space containers (physical) for all types of table spaces.
    • Perform the direct reads.
    • Read data and index pages from the table space containers (physical) by asynchronous engine dispatchable units (EDUs) for all types of table spaces. Refer to the DB2 documentation for the following monitor elements:
  • Write time: Average time per second spent to:
  • Total wait time: Average time spent to waiting within DB2 per second. Refer to the DB2 documentation for the following monitor elements:

MySQL

  • Active session count: Number of currently active user sessions, excluding sleeping sessions and background processes.
  • All session count: Number of currently established user sessions, including sleeping ones, but excluding background processes.
  • Delete statements per second: Average number of DELETE statements requested per second. Note that this value correspond to the number of request issued and might not be corresponding to number of requests successfully completed.
  • Insert statements per second: Average number of INSERT statements requested per second. Note that this value correspond to the number of request issued and might not be corresponding to number of requests successfully completed.
  • Select statements per second: Average number of SELECT statements requested per second. Note that this value correspond to the number of request issued and might not be corresponding to number of requests successfully completed.
  • Update statements per second: Average number of UPDATE statements requested per second. Note that this value correspond to the number of request issued and might not be corresponding to number of requests successfully completed.
  • Executed queries per second: Average number of statements executed per second by the database server. This includes only statements sent to the server by clients and not statements executed within stored programs. This variable does not count COM_PING, COM_STATISTICS, COM_STMT_PREPARE, COM_STMT_CLOSE, or COM_STMT_RESET commands.
  • Number of bytes read from files: Average number of bytes read per second from all files.
  • Number of bytes written to files: Average number of bytes written per second to all files.
  • Table cache hit rate: Average cache hit ratio, in percent.
  • Total wait time per second: Average time spent waiting on events per second. Idle events are ignored.

Oracle and Oracle RAC

  • Active session count: Number of active, non-idle sessions currently executing SQLs.
  • All session count: Number of all sessions, including active, inactive, idle, and others.
  • Average physical read bytes: Average number (in bytes per second) of disk reads by all database instance activity, including application reads, backup and recovery, and other utilities.
  • Average physical write bytes: Average number (in bytes per second) of disk writes for all database instance, including application activity, backup and recovery, and other utilities.
  • Configured nodes (RAC only): Number of nodes configured in cluster, reported by each database instance (node), including offline or currently not responding nodes. Configured nodes mean nodes, that are part of this cluster database, even if they are offline or malfunctioning at the moment.
  • Connected nodes (RAC only): number of connected (online, active and working) cluster nodes, reported by each database instance (node). Connected nodes mean nodes, that are currently online, active and working.
  • SQL parse time: Average time spent on parsing SQL statements per second.
  • SQL execute time: Average time spent on executing SQL statements per second. Note that this also includes the amount of time spent performing fetches of query results, but does not include network latency outside the instance, which is part of actual end-to-end response time for a SQL statement.
  • PL/SQL execute elapsed time: Average time spent running the PL/SQL interpreter per second. This does not include time spent recursively executing/parsing SQL statements or time spent recursively executing the Java VM.
  • Connection management: Average number of seconds spent on connection management per second, including time spent performing session connect and disconnect calls.
  • Time spent on other activities: Average number of seconds spent on other activities per second. This includes time spent performing Database user-level calls, excluding instance background processes and four other activities measured by other metrics (SQL parse time, SQL execute time, PL/SQL execute elapsed time, Connection management).
  • Total wait time per second: Average time spent by all processes waiting for events per second. This value is regarding instance-wide time for all non-idle wait classes (which is Application, Configuration, Administrative, Concurrency, Commit, Network, User I/O, System I/O, Scheduler and Other).
  • User calls count: Average number of user calls (such as login, parse, fetch, or execute) per second.
  • CPU idle: Current CPU idle cycles, in percent. CPU is described as idle when it is not being used by any process.
  • Foreground CPU usage: Current CPU load caused by the foreground database processes, in percent. A change in the metric value may occur because of a change in either workload mix or workload throughput being performed by the database. Although there is no ‘correct’ value for this metric, it can be used to detect a change in the operation of a system.
  • Background CPU usage: Current CPU load caused by the background database processes, in percent. A change in the metric value may occur because of a change in either workload mix or workload throughput being performed by the database. Although there is no ‘correct’ value for this metric, it can be used to detect a change in the operation of a system.
  • Non Database CPU usage: Current CPU load caused by non-database operating system processes, in percent. A change in the metric value may occur because of other processes executed on same host, including OS-specific processes. Although there is no ‘correct’ value for this metric, it can be used to detect a change in the operation of a system.

SQL Server

  • All sessions count: Number of all authenticated sessions (including running, sleeping, dormant, and preconnected).
  • Running Session Count: Number of authenticated sessions currently running one or more requests.
  • Batch requests per second: Number of T-SQL command batches received by the server per second.
  • Database CPU idle: Current CPU idle cycles, in percent. CPU is described as idle when it is not being used by any process.
  • Database CPU usage: Current CPU load caused by the database processes, in percent. A change in the metric value may occur because of a change in either workload mix or workload throughput being performed by the database. Although there is no ‘correct’ value for this metric, it can be used to detect a change in the operation of a system.
  • Non Database CPU usage: Current CPU load caused by non-database operating system processes, in percent. A change in the metric value may occur because of other processes executed on same host, including OS-specific processes. Although there is no ‘correct’ value for this metric, it can be used to detect a change in the operation of a system.
  • Number of bytes read per second issued on the files: Number of bytes read per second, for all files in all databases in the SQL Server instance.
  • Number of bytes written per second issued on the files: Number of bytes written per second, for all files in all databases in the SQL Server instance.
  • SQL Compilations per sec: Number of query and procedure compilations per second.
  • SQL Re-Compilations per sec: Number of query and procedure recompilations per second.
  • Total wait time per second: Average time spent by all processes waiting on events per second. This includes waits encountered by threads that executed.