Reaction configurations allow you to customize various mechanisms controlling how reactions behave. The following parameters are available to the zz$manage.reaction_config procedure:
|ALERT_PREFIX||String, recommended is a specific Oracle Error Code, such as ORA-20899.||Prefix to prepend to all messages written to the alert log, if the alert log writing reaction is configured.|
|AUD_LIMIT||Numeric, set to limit the number of rows processed in a single bulk operation.||This parameter controls how much memory is used when processing records in aud$. Only valid if the zztat Auditing Add-on has been licensed. Defaults to 10000.|
|AUTO_PURGE_INTERVAL||Number of days.||Controls the retention of metric data on the repository. Data older than the specified number of days will be automatically purged. This also directly influences space requirements of the repository.|
|CRIT_EMAIL||String, email address.||Email address that is to receive emails sent by the SEND_EMAIL reaction. Used for critical alerts.|
|DATAFILE_ADD_LOCATION||String, either an ASM diskgroup name incl the plus sign (for example +DATA), or the fixed value "USE_DB_CREATE_FILE_DEST".||When this is set, automatic data file addition can be enabled. If this is set to an ASM diskgroup name, that diskgroup is used as long as there is sufficient space available.
When this is set to USE_DB_CREATE_FILE_DEST, zztat will retrieve the value of the init.ora parameter db_create_file_dest and use that. db_create_file_dest must be set to an ASM diskgroup and cannot be set to a directory path. Automatic data file addition on file systems is currently unsupported.
|DATAFILE_ADD_IF_GB_FREE||Numeric, minimum free space threshold in gigabytes.||Set to the minimum usable free space in the ASM diskgroup to still allow data files to be added automatically. If usable free space falls below this threshold, files will not be added.|
|DATAFILE_ADD_COMMAND||String, file size specification to use when adding datafiles.||The default value is "size 1g autoextend on next 1g maxsize unlimited". Any valid Oracle syntax can be used. Note: zztat does not validate / parse this string. It is passed to the alter tablespace command as specified here.|
|DEFAULT_REACTION||Specify the name of an existing reaction.||The reaction specified here is used as a fallback, when the reaction that was triggered by the gauge fails.|
|HEARTBEAT_TIMEOUT||Number of seconds.||If the repository database has not received a ping from a target database after this many seconds have elapsed, it will cause an alert to be raised.|
|HEARTBEAT_TASK_INTERVAL||Number of seconds.||Specifies the number of seconds between a target's checks with the repository to see if there are any tasks for it to perform. This directly controls the responsiveness of the framework as a whole. Lower values cause changes done on the repository database to be propagated faster to the target databases. Valid values are between 10 and 300 seconds (5 minutes).
Note: when deploying a zztat patch, the framework will automatically lower this value to 10 seconds, and restore the previous value when patching completes.
|LATCH_SAMPLES||Numeric, number of samples to take||This controls the high-speed latch reaction LATCH_HISPEED_SAMPLE and governs the number of samples taken. Defaults to 1000 which will take less than a second on modern CPU systems. Higher values may yield more exact measurements - 10000 or 100000 are good values to use when specific latch contention monitoring is desired.|
|LOG_DIRECTORY||String, must point to an existing directory on the database server. Can include a trailing slash, but does not need to.||Directory object to use for the WRITE_FILE reaction, where the LOG_FILE value will be created.|
|LOG_FILE||String, a file name that will be created in the LOG_DIRECTORY.||Logfile to use for the WRITE_FILE reaction. Placeholders, such as ##DATE## will be added in a future release.|
|LOG_TABLE||SCHEMA.TABLE_NAME or just TABLE_NAME.||Custom log table to write errors to using the WRITE_TABLE reaction.|
|MARKED_SQL_AGE_LIMIT||Number of days||Marked SQL will be automatically aged out of the staging areas after this # of days since it was last seen.|
|MUTEX_SAMPLES||Numeric, number of samples to take||This controls the high-speed mutex reaction MUTEX_HISPEED_SAMPLE and governs the number of samples taken. Defaults to 1000 which will take less than a second on modern CPU systems. Higher values may yield more exact measurements - 10000 or 100000 are good values to use when specific mutex monitoring is desired.|
|OTHER_SAMPLES||Numeric, number of samples to take||This controls the non-latch and non-mutex hi-speed sampling, such as x$kslwt for hi-speed session wait sampling.|
|SEND_ERROR_REPORT||Specify either YES or NO.||When enabled, this causes the framework to send an error report email to the zztat developer. This allows us to proactively react to framework errors and / or programmatic errors that happen on zztat environments. These errors reports do not contain any information that could be used to identify the environment, such as IP addresses or host names. See this post for more details.|
|TARGET_AUTO_PATCH||Specify either YES or NO.||When enabled, causes a target database to automatically synchronize its patch level with the repository database and transfer and apply any patches it is missing. Targets will not automatically roll back patches - this only works for apply.|
|WARN_EMAIL||String, email address.||Email address that is to receive emails sent by the SEND_EMAIL reaction. Used for warnings.|