SYS.MONITOR

The MONITOR table stores information about system performance. It contains a single row with statistics about certain events. For many columns, statistics are gathered starting from the time the database is loaded into memory and statistics are cleared when the database is unloaded from memory. With a ramPolicy of manual or always, the database remains in memory when there are no application connections.

For some columns, statistics are gathered as needed. TimesTen does not gather statistics from the time of the first connection for these columns:

PERM_ALLOCATED_SIZE
PERM_IN_USE_SIZE
TEMP_ALLOCATED_SIZE
LAST_LOG_FILE
REPHOLD_LOG_FILE
REPHOLD_LOG_OFF
FIRST_LOG_FILE
CHECKPOINT_BYTES_WRITTEN

For most columns, the MONITOR table is reset whenever there are no connections to the database. TimesTen does not reset the values of the following columns, even when there are no connections to the database:

PERM_ALLOCATED_SIZE
PERM_IN_USE_SIZE
TEMP_ALLOCATED_SIZE
LAST_LOG_FILE
REPHOLD_LOG_FILE
REPHOLD_LOG_OFF
FIRST_LOG_FILE

TimesTen frequently updates information in the MONITOR table. To prevent these updates from slowing down the system, they are not protected by latches. Hence values in the MONITOR table are not absolutely accurate. They can be used as a reliable indication of activities in the system.

Note:

See the SYS.SYSTEMSTATS table if the desired statistic is not in the SYS.MONITOR table. There are statistics that exist in both the SYS.MONITOR table and in the SYS.SYSTEMSTATS table.

Columns

Column name	Type	Description
`TIME_OF_1ST_CONNECT`	`TT_CHAR(32) NOT NULL`	Time at which the first connection was made
`DS_CONNECTS`	`TT_INTEGER NOT NULL`	Number of connects to the database
`DS_DISCONNECTS`	`TT_INTEGER NOT NULL`	Number of disconnects from the database
`DS_CHECKPOINTS`	`TT_INTEGER NOT NULL`	Number of checkpoints taken
`DS_CHECKPOINTS_FUZZY`	`TT_INTEGER NOT NULL`	Number of fuzzy checkpoints taken
`DS_COMPACTS`	`TT_INTEGER NOT NULL`	Number of database compactions
`PERM_ALLOCATED_SIZE`	`TT_BIGINT NOT NULL`	Size in kilobytes of the memory allocated to the permanent region
`PERM_IN_USE_SIZE`	`TT_BIGINT NOT NULL`	Size in kilobytes of the memory in the permanent region that is currently in use
`PERM_IN_USE_HIGH_WATER`	`TT_BIGINT NOT NULL`	The highest amount (in kilobytes) of memory in use within the permanent region since the first connection to the database The value of this field can be reset to the current value of the `PERM_IN_USE_SIZE` attribute by using the `ttMonitorHighWaterReset` built-in procedure.
`TEMP_ALLOCATED_SIZE`	`TT_BIGINT NOT NULL`	Size in kilobytes of the memory allocated to the temporary region
`TEMP_IN_USE_SIZE`	`TT_BIGINT NOT NULL`	Size in kilobytes of the memory in the temporary region that is currently in use
`TEMP_IN_USE_HIGH_WATER`	`TT_BIGINT NOT NULL`	The highest amount (in kilobytes) of memory in use within the temporary region since the first connection to the database The value of this field can be reset to the current value of the `TEMP_IN_USE_SIZE` attribute by using the `ttMonitorHighWaterReset` built-in procedure.
`TPL_FETCHES`	`TT_BIGINT NOT NULL`	Number of times TimesTen fetches data from the Oracle database into TimesTen using dynamic load
`TPL_EXECS`	`TT_BIGINT NOT NULL`	Number of times TimesTen communicates with the Oracle database to dynamically load data into TimesTen This count also includes attempts to perform a dynamic load when there is no data to fetch from the Oracle database.
`CACHE_HITS`	`TT_BIGINT NOT NULL`	Number of times TimesTen successfully finds the required data in TimesTen
`PASSTHROUGH_COUNT`	`TT_BIGINT NOT NULL`	Number of successful passthrough executions
`XACT_BEGINS`	`TT_BIGINT NOT NULL`	Number of transactions started
`XACT_COMMITS`	`TT_BIGINT NOT NULL`	Number of durable and nondurable transactions committed
`XACT_D_COMMITS`	`TT_BIGINT NOT NULL`	Number of transactions committed durably
`XACT_ROLLBACKS`	`TT_BIGINT NOT NULL`	Number of transactions rolled back
`LOG_FORCES`	`TT_BIGINT NOT NULL`	Number of times log files were synchronized to disk
`DEADLOCKS`	`TT_BIGINT NOT NULL`	Number of deadlocks
`LOCK_TIMEOUTS`	`TT_BIGINT NOT NULL`	Number of lock requests denied due to timeouts
`LOCK_GRANTS_IMMED`	`TT_BIGINT NOT NULL`	Number of lock requests granted without a wait
`LOCK_GRANTS_WAIT`	`TT_BIGINT NOT NULL`	Number of lock requests granted after a wait
`CMD_PREPARES`	`TT_BIGINT NOT NULL`	Number of commands prepared (compiled)
`CMD_REPREPARES`	`TT_BIGINT NOT NULL`	Number of commands re-prepared
`CMD_TEMP_INDEXES`	`TT_BIGINT NOT NULL`	Number of temporary indexes created during query execution
`LAST_LOG_FILE`	`TT_INTEGER NOT NULL`	Most recent log file present Same as `log.file.latest` in `SYS.SYSTEMSTATS`.
`REPHOLD_LOG_FILE`	`TT_INTEGER NOT NULL`	Number of last log file held by replication
`REPHOLD_LOG_OFF`	`TT_INTEGER NOT NULL`	Offset in last log file held by replication
`REP_XACT_COUNT`	`TT_INTEGER NOT NULL`	The number of replicated transactions generated on the local database that are being replicated to at least one peer database
`REP_CONFLICT_COUNT`	`TT_INTEGER NOT NULL`	The number of replicated transactions that ran into a conflict when being applied on the local database
`REP_PEER_CONNECTIONS`	`TT_INTEGER NOT NULL`	The sum of all peer connections initiated by the local replication agent There is one connection for every peer relationship where the local database is the master. If a transport level failure results in the establishment of a new connection, this count is incremented.
`REP_PEER_RETRIES`	`TT_INTEGER NOT NULL`	The number of retry attempts while trying to establish a new peer connection
`FIRST_LOG_FILE`	`TT_INTEGER NOT NULL`	The number of the oldest existing (not yet purged) log file
`LOG_BYTES_TO_LOG_BUFFER`	`TT_BIGINT NOT NULL`	The number of bytes written to the log since first connect This value includes the sizes of actual log records plus any log overhead.
`LOG_FS_READS`	`TT_BIGINT NOT NULL`	The number of times that a log read could not be satisfied from the in-memory log buffer
`LOG_FS_WRITES`	`TT_BIGINT NOT NULL`	The number of times TimesTen has written the contents of the in-memory log buffer to the operating system This column does not count the number of times data was flushed to disk. It counts writes to the operating system's file buffers.
`LOG_BUFFER_WAITS`	`TT_BIGINT NOT NULL`	The number of times a thread was delayed while trying to insert a log record into the log buffer because the log buffer was full Generally speaking, if this value is increasing, it indicates that the log buffer is too small.
`CHECKPOINT_BYTES_WRITTEN`	`TT_BIGINT NOT NULL`	The number of bytes written to disk by the most recent checkpoint operation
`CURSOR_OPENS`	`TT_BIGINT NOT NULL`	Number of `SELECT` statements issued
`CURSOR_CLOSES`	`TT_BIGINT NOT NULL`	Number of `SELECT` statements completed
`CHECKPOINT_BLOCKS_WRITTEN`	`TT_BIGINT NOT NULL`	Total number of blocks written for all completed checkpoints To improve I/O efficiency, multiple blocks may be coalesced into a single write or a single block may be split across multiple writes.
`CHECKPOINT_WRITES`	`TT_BIGINT NOT NULL`	Total number of write operations issued by all completed and in-progress checkpoints
`REQUIRED_RECOVERY`	`TT_INTEGER NOT NULL`	Recovery status 1: When the database was initially loaded into RAM at `TIME_OF_1ST_CONNECT`, recovery ran. This means that the previous time the database was in memory, the database did not shut down cleanly. When it was loaded into memory this time, the log was replayed and other operations were performed in an attempt to recover data. If `DurableCommit` had been set to 0, transactions could have been lost. 0: The database was previously shut down cleanly. Thus the database was restarted cleanly.
`TYPE_MODE`	`TT_INTEGER NOT NULL`	0: Oracle mode