This documentation is for an older version of HVR.
Regular Maintenance and Monitoring for HVR
Periodically it is necessary to restart the HVR Scheduler, purge old log files and check for errors. HVR must also be monitored in case a runtime error occurs.
Both maintenance and monitoring can be performed by script Hvrmaint. This is a standard Perl script which does nightly or weekly housekeeping of the HVR Scheduler. It can be scheduled on Unix or Linux using crontab or on Windows as a Scheduled Task.
HVR can also be watched for runtime errors by an enterprise monitoring system, such as TIVOLI or UNICENTER. Such monitoring systems complement Hvrmaint instead of overlapping it. Such systems can watch HVR in three ways:
- Check that the Hvrscheduler process is running in the operating system process table.
- Check that no errors occur in file $HVR_CONFIG/log/hubdb/hvr.crit or in file $HVR_ITO_LOG (see section Hvrscheduler).
- Check that file $HVR_CONFIG/log/hubdb/hvr.out is growing.