CHECKPOINT is an executable command that can be issued interactively or embedded in a host language.
One of the factors affecting the length of time needed to return location data objects to a logically consistent state after a transient system failure (for example, power outage, operating system failure) is the value of the lCheckpointFreq configuration parameter. Changing the value of this parameter (set in the [XDBLogManager] section of thexdbsrvco.ini file) allows the user to increase or decrease the time interval between automatic checkpoints issued by the logging system.
lCheckpointFreq is defaulted to a value of 300 seconds, which means that the logging system automatically performs a checkpoint every five minutes. If this checkpoint interval is adjusted upward, more time could be necessary for log recovery after a transient system failure. Conversely, a shorter checkpoint interval could reduce average recovery time, but increase system overhead. See the Server Administration Guide for more information on the lCheckpointFreq parameter.
Description
Under normal operating conditions CHECKPOINT commands are issued automatically by the logging system at regular time intervals (see lCheckpointFreq Parameter below). Some occasions, however, warrant issuing a CHECKPOINT command within SQLWizard as a precautionary measure. If you know (or suspect) that the server must be taken down, and you can still access SQLWizard, then issue a CHECKPOINT. This action minimizes crash recovery time when the server is brought back up.