How to avoid fragmentation
Java SDK Release 1.3.1 at SR7, and later, provides command-line options to specify the size of the JVM kCluster and pCluster regions. Refer to Table | -4 for a summary of the switches. For additional information see the IBM Java technology Web page "Diagnosis Documentation," which has a section on garbage collection and performance tuning. The Web address is:
http://www-128.ibm.com/developerworks/java/jdk/diagnosis/142.html
Table 2-4
| JVM
region | JVM
switch | |
|---|---|---|
| kCluster | -Xk | |
| pCluster overflowsize | -Xp sz,ovfl | The -Xp switch has two parameters: sz = the pCluster size parameter in KB ovfl = the overflow size parameter in KB |
JVM kCluster and pCluster sizing switches
Set the initial sizes of the clusters large enough to help avoid fragmentation issues occurring on your Web site. It is not unusual in a large Java application, such as WebSphere Application Server, that the default kCluster space might not be sufficient to allocate all class blocks.
Example 2-18 Sample garbage collection output using -verbosegc switch
<GC(VFY-SUM): pinned=4265(classes=3955/freeclasses=0) dosed=10388 movable=1233792 free=5658>
In Example | -18, the pinned size value(4265) and classes size value(3955) are about the right size needed for the -Xk parameter. However, we recommend that you add 10% to the reported value (3955)
In the preceding example, -Xk4200 is probably a reasonable setting. The difference between pinned (=4265) and classes (=3955) provides a guide for the initial size of pCluster; however, because each object might be a different size, it is difficult to predict the requirements for the pCluster and pCluster overflow options.