The following is information on IBM supplied predefined performance monitors. Each IBM supplied predefined monitor is listed below with a brief description of what it can be used for. The only attribute of these monitors that can be changed is the Default for level attribute. When you are viewing a list of monitors, NULLID appears in the Created by column for all IBM supplied predefiend monitors.
Monitoring Capacity
Use this monitor to get information on system capacity. For example, if you notice that the system is slowing down over time, the number of connections might have been increasing. This monitor can be checked on a regular basis to see the overall usage of your system over time.
Sort
Use this monitor to ensure that your sort heap and sort heap threshold parameters are set correctly. Running this monitor will tell you if your sort heap is set too high or too low, how much sorting is going on, and if you need to increase your sort heap threshold parameter. This monitor should be run when you first start your system, in peak periods of activity, or as applications change.
Locking
Use this monitor to determine how much locking is occurring in your system, and whether your lock list parameters are set appropriately. In a Data Warehousing environment, you might not see many lock wait situations, therefore, you might not need to use this monitor.
Cache
Use this monitor to optimize cache usage. By monitoring these values during peak periods, you can determine if you need to increase your cache size.
Bufferpool
Use this monitor on small tables to determine whether they require their own bufferpools. In a data warehousing environment, this monitor might not be very useful because it is unlikely that you will have a high bufferpool hit ratio on your fact table.
Note: The greater the bufferpool hit ratio, the lower the frequency of disk input and output.
Deadlocks
Use this monitor to determine whether your applications are getting into deadlocks. (This is very unlikely for data warehousing workloads.) If the total number of deadlocks grows at all, examine the database configuration parameters DLCHKTIME and LOCKTIMEOUT and turn on the event monitor for deadlocks to collect information on the transactions involved in the deadlock.
Fast Communication Manager
The Fast Communication Manager buffer is the memory used to transfer information between nodes. If the minimum values become low or if the value drops below 10% free, then you should increase the associated database manager configuration parameters.
Prefetchers
This monitor monitors the prefetch satisfied variable. Use this monitor during peak load to determine whether you have enough prefetchers defined for the system. You should also examine the database configuration parameter DFT_PREFETCHSZ.
Disk Performance
Use this monitor to monitor input and output. This monitor contains performance variables that focus on disk performance at the Database and Tablespace level.
Global Memory
Use this monitor to monitor application memory use.
Long Running Memory
Use this monitor to help determine why a query is taking a long time to complete.
Gateway Connections
Use this monitor to monitor gateway connections.
Default monitor for the Database level
Use this monitor for general Database tuning.
Default monitor for the Table level
Use this monitor for general Table tuning.
Default monitor for the Instance level
Use this monitor for general Instance tuning.
Default monitor for the Tablespace level
Use this monitor for general Tablespace tuning.
Default monitor for the Connections level
Use this monitor for general Connections tuning.