WebSphere eXtreme Scale Best Practices for Operation and Management

122 WebSphere eXtreme Scale Best Practices for Operation and Management

Example 6-7 JVM output messages when the memory threshold default value is changed

CWOBJ0046I: The memoryThresholdPercentage property was not provided in a server

properties file and the MemoryMXPool bean was not set (current value = 0). A

default value of 70 will be used.

CWOBJ0034I: Memory utilization threshold percentage is set to 70 %.

CWOBJ0036W: Changing memory utilization threshold from 0K to 367001K for Java

heap memory pool.

CWOBJ0037W: Changing memory collection utilization threshold from 0K to 367001K

for Java heap memory pool.

CWOBJ0034I: Memory utilization threshold percentage is set to 60 %.

CWOBJ0036W: Changing memory utilization threshold from 367001K to 314572K for

Java heap memory pool.

CWOBJ0037W: Changing memory collection utilization threshold from 367001K to

314572K for Java heap memory pool.

CWOBJ0034I: Memory utilization threshold percentage is set to 60 %.

6.3 Locking performance preferred practices

An ObjectGrid BackingMap supports several locking strategies to maintain cache entry

consistency. You can configure each BackingMap to use one of the following locking

strategies:

򐂰 Pessimistic locking mode

򐂰 Optimistic locking mode

򐂰 None

Set the locking strategy on the BackingMap. You cannot change the locking strategy for each

transaction. The following example shows how to set the locking mode on a map using the

XML file:

<backingMap name="myMap" lockStrategy="PESSIMISTIC" />

6.3.1 Pessimistic locking strategy

The pessimistic locking strategy acquires locks for cache entries. Use this strategy when data

is changed frequently. Any time that a cache entry is read, a lock is acquired and held

conditionally until the transaction completes. The duration of certain locks can be tuned using

transaction isolation levels for the session.

6.3.2 Optimistic locking strategy

Optimistic locking is the default configuration. This strategy improves both performance and

scalability when compared to the pessimistic strategy. Use this strategy when your

applications can tolerate occasional optimistic update failures. They perform better than they

will by using the pessimistic strategy. This strategy is excellent for read operations and

infrequent update applications.

Messages: The first message showing that the memoryThresholdPercentage was not

provided is quite confusing. It is obviously contradicted with the next message showing that

the memory threshold percentage was set to another value than the default. These

messages are normal in the current product.