Plan your fix management strategy

Start of change Use this information to establish a maintenance strategy to reduce the impact to i5/OS™ operations that result from unplanned outages or program failures. End of change

To most people, preventive maintenance means changing the oil in your car Start of change every 3000 miles, rotating the tires every 10 000 miles, and getting major maintenance done every 30 000 miles. End of change This is the standard maintenance strategy automobile manufacturers suggest for new vehicles to prevent expensive problems.

IBM has similar guidelines to help you develop an effective program maintenance strategy. These guidelines are intended to provide basic program maintenance definitions, information, and direction for new users or for those who currently do not have a program maintenance strategy in place.

Why an i5/OS maintenance strategy is recommended

Three out of four defect-related problems that are reported are rediscoveries of previously reported problems. Many users may have avoided the problem or outage if the available fix had been applied to their system. Unplanned outages have a tremendous impact on employee productivity, business operations, and revenue.

Recommendations for setting up an effective maintenance strategy

Start of change Unfortunately, there is no single recommendation for fix maintenance. Each system or environment must be assessed individually. Use the Fix Maintenance Advisor to help you determine a strategy for preventive maintenance that is based on your iSeries™ environment and applications. As you develop your strategy, here are some questions that you need to consider: End of change

What are you doing to prevent unexpected failures associated with i5/OS licensed programs, including interruptions to communications networks or unscheduled outages on your iSeries servers?
Is your standard approach to program maintenance reactive, that is, you apply corrective fixes when failures occur?
Do you have a preventive maintenance strategy in place for your iSeries servers?
Is your system in a 24x7 production environment that requires maximum availability, or is it limited to testing new applications and used only during prime shift Monday through Friday by a limited set of programmers?
Is your server on a new software release or on a release that has proven stable in your environment?
What would be the tolerance and cost to the business of an unexpected system outage?

Start of change Your preventive maintenance schedule might resemble the following tasks: End of change

Time frame	Tasks
Daily	Run backup procedure Apply individual fixes as needed
Weekly	Order and review Preventive Service Planning (PSP) Order and apply necessary HIPER PTFs Print and review hardware error log (PRTERRLOG)
Monthly for a changing server or quarterly for a stable server	Order current cumulative PTF packages (SF99vrm) or PTF Group packages (SF99nnn) and apply Do a complete system save Perform system cleanup functions
Notes: In a stable environment, think about installing the most recent cumulative PTF package every three to four months. Consider installing the most recent cumulative PTF package before making major changes to your system (hardware or software). If you require an excessive number of corrective service fixes, consider installing cumulative PTF packages more frequently.

When you establish a maintenance strategy, you can reduce the effect on the i5/OS operations that result from unplanned outages or program failures. When you use a program maintenance strategy that is appropriate for your environment, you can provide optimal system performance and availability.