Business as Usual 1
- Jason Andrew
Once the Instance is up and running, there are a number of regular tasks that must be undertaken to ensure the continued, healthy running of that instance. Remember, the Administrators are providing 1st and 2nd line Support for the system, so it is advisable to proactively prevent possible outage situations occurring. Consider the following areas:
Tidying of Logfiles
Are the logfiles being managed?
Many of the processes create log files and, if not tidied, these could cause the instance to run out of space. In the extreme this can possibly cause the system to crash. Filestore HouseKeeping gives pointers on managing this.
Tidying the data
Is the historic data being managed?
As jobs are booked, and user interactions occur within the instance, a number of new table entries are created which, overtime, can cause the database to become clogged with old data that might affect performance - this is particularly true if the auditing features of the product are enabled. Several product features enable the tidying of this data in a managed fashion - see Data HouseKeeping for further details.
Checking for failed processes
The business may not have complained of any outages - but has any part of the system restarted during the day?
Since the ServiceOptimizer system is capable of automatically restarting the individual components, should they crash, it is possible that a crash has occurred that the business users have not been aware of. Whilst these crashes are not causing an immediate major outage it is worth following up on the reasons for them as they might be precursors to a more major incident. Any failed processes should be investigated and, if necessary, reported to ServicePower Support. System Configuration gives details of which process one would expect to see running within the healthy system, and Application Logfiles gives details of the log files produced when a process restarts. These sections should help with how to go about with the investigation process.
Regular restarts of any part of the instance
Is the instance set up correctly for component restart?
Planned Restart of Components covers how restarts may be configured. It is useful to reconsider these settings in the light of any changes to business practice, reorganisation of geography etc .
Checking Database Performance
Are there any indications of bottlenecks within the database?
Additionally, the database functionality itself needs to be regularly reviewed. It is assumed that a DBA will be managing the instance and will take appropriate steps with regards to housekeeping the DB - rebuilding indexes etc on a regular basis. Oracle HouseKeeping gives further details.
Checking Server Performance/Resources
Are there any indications of potential issues on the performance front?
The mechanisms for determining this will vary from O/S to O/S but things that need to be considered are:
- Number of file handles being used
- Number of ports in use
- Turn around time for API calls
Keeping a regular eye on these aspects of the system may help identify when a bottle neck has occurred and give early warning before a potential outage occurs as a result. Refer to Resources for more detail.
Checking Business Requirements
Have any business requirements changed since the system was set up?
As the business runs many things can change that might affect the healthy running of the Instance. It is a good idea to have regular reviews with the business to ensure that the set up of the instance does not need altering as a result. Things to consider here are:
- Large numbers of new employees have been added or are planned to be added - which FRUs will they go in?
- The business may have changed their process flow
Refer to System Configuration to see what may need to be changed as a result.