System maintenance

Once a PME system is installed and commissioned, it must be maintained on a regular basis. Without periodic, proactive system maintenance, system performance degrades over time and the system operates with poor performance (timeouts and sluggish response) and eventually becomes unusable.

This section provides guidelines and recommendations for creating a maintenance schedule. Proper adherence to the maintenance guide ensures a steady-state operation of a PME system.

This section does not detail troubleshooting system issues, but identifies the steps to develop a maintenance plan that should allow for a commissioned system to continuously run at optimal performance.

A maintenance plan keeps your PME system healthy. A comprehensive maintenance plan starts in the system planning stage before commissioning. The plan should include frequent monitoring and system reviews, and maintenance tasks to ensure the system is operating under healthy conditions.

The following sections provide recommendations for developing your maintenance plan:

• Designing for maintenance

• Database maintenance

• System health review

Designing for maintenance

Before installing and deploying a system, it is important to plan and design the IT architecture in support of maintenance and backup activities. Maintenance and backup activities often require additional resources (CPU, RAM and hard drive space) to perform and complete the activity. The recommendations are as follows:

• System sizing

  Size a system for the final state (for example: maximum devices, maximum logging parameters, etc) to ensure that the system runs in a healthy state for a longer time. See IT Requirements for recommended system sizing.

• Storage space allocation

  Make sure there is enough hard drive space to perform database maintenance, archive and trim operations. See Storage Performance and Availability for recommended hard drive sizing.

• Disk I/O performance

  Disk read/write operations impacts system performance. Plan to have different storage devices installed for the system files, the database files, and the Microsoft Message Queuing (MSMQ) files. See Choosing Data Storage for recommended storage information.

• Pre-size databases

  By default, ION_Data database is configured with a 10% auto-growth parameter. This growth operation can lead to a fragmented database and hard drive and can impact database performance over time. To minimize the use of the auto-growth feature, pre-allocate hard drive space for the ION_Data database during system commissioning, including systems running SQL Server Express. See Database growth for recommendation on database growth.

• Database auto growth

  For systems expected to have large ION_Data databases (> 100 GB), change the default auto-growth setting to a fixed size rather than by percentage (Default is 10%). In this case, the database auto-grows when the database exceeds the preset container size. This ensures better control of database growth. Auto-growth events are expensive operations that slow down the performance of your database. Take the following preventive steps to ensure your auto-growth is in control:

  • Minimize future auto-growth events by monitoring the growth of your databases, and re-establishing auto-growth settings when a database growth profile changes.

  • Monitor auto-growth events so you can be alerted when your databases grow.

  • Consider defragmenting your database file system.

  See SQL Server Database Growth and Autogrowth Settings for more information.