Tuning parameter hot list
WebSphere® Application Server provides several tunable parameters and options to match the application server environment to the requirements of your application.
- Review the hardware and software requirements
It is critical for proper functionality and performance to satisfy the minimum hardware and software requirements. Refer to IBM® WebSphere Application Server supported hardware, software, and APIs website which details hardware and software requirements.
- Install the most current refresh pack, fix pack, and the recommended interim
fixes
The list of recommended updates is maintained on the Support site.
- Check hardware configuration and settings
Check network connections to make sure that they are running at their highest speed. For more information, see Tuning the application serving environment.
Verify network interconnections and hardware configuration is setup for peak performance.
- Tune the operating
systems
Operating system configuration plays a key role in performance. For example, adjustments such as TCP/IP parameters might be necessary for your application
- Tuning IBM i systems
Operating system configuration plays a key role in performance. For example, adjustments such as TCP/IP parameters might be necessary for your application.
- Set the minimum and maximum Java™ virtual machine (JVM) heap
sizes
Many applications need a larger heap size then the default for best performance. It is also advised to select an appropriate GC policy based on the application's characteristics.
- Use a type 4 (or pure Java) JDBC driver
In general, the type 2 JDBC driver is recommended if the database exists on the same physical machine as the WebSphere instance. However, in the case where the database is in a different tier, the type 4 JDBC driver offers the fastest performance since they are pure Java not requiring native implementation. Use the previous link to view a list of database vendor-specific requirements, which can tell you if a type 4 JDBC driver is supported for your database.
- Use a type 2 JDBC driver for local data access
and a type 4 (or pure Java) JDBC driver for remote data access
In general, the type 2 JDBC driver is recommended. Use the previous link to view a list of database vendor-specific requirements, which can tell you if a type 4 JDBC driver is supported for your database.
- Tune WebSphere Application Server JDBC data sources and associated connection
pools
The JDBC data source configuration might have a significant performance impact. For example, the connection pool size and prepared statement cache need to be sized based on the number of concurrent requests being processed and the design of the application.
See the Connection pooling topic for more information.
- Enable the pass by reference
option
Use applications that can take advantage of the pass by reference option to avoid the cost of copying parameters to the stack.
- Ensure that the
transaction log is assigned to a fast disk
Some applications generate a high rate of writes to the transaction log. Locating the transaction log on a fast disk or disk array can improve response time.
- Tune related components, such as a database
In many cases, some other component, for example, a database, needs adjustments to achieve higher throughput for your entire configuration.
- Disable functions that are not required
For example, if your application does not use the web services addressing (WS-Addressing) support, disabling this function can improve performance.
Attention: Use this property with care because applications might require WS-Addressing MAPs to function correctly. Setting this property also disables WebSphere Application Server support for the following specifications, which depend on the WS-Addressing support: Web Services Atomic Transactions, Web Services Business Agreement and Web Services Notification.
To disable the support for WS-Addressing, refer to Enabling Web Services Addressing support for JAX-RPC applications.
- Tuning index
One of the goals of the product programming model and runtime is to significantly simplify the work required for application developers to write and deploy applications. Sometimes we say that the product relieves the application programmer of many of the plumbing tasks involved in developing applications. For example, application code in the product does not concern itself directly with remote communication--it locates objects which may be local or remote and drives methods. Therefore, you won't see any direct use of socket calls or TCP/IP programming in this application code.
This separation of what you want to do from where you do it is one aspect of removing the application programmers from plumbing tasks. Other considerations are not having to deal with data calls for some types of beans, potentially user authentication, and threading. There are generally no calls from the application code to touch sockets, RACF® calls, or management of threading. Removing this from the application programmer doesn't mean this work won't get done. Rather, it means that there may be more work for the DBA, the network administrator, the security administrator, and the performance analyst.
There are four layers of tuning that need to be addressed:We deal with the first three in separate sections under this article and briefly touch on the fourth. For more information on tuning applications, refer to Developing client applications.
- Tuning the subsystems
Steps involved in tuning the z/OS subsystems to optimize product performance include:
- Tuning the product runtimeSteps involved in tuning the product runtime to optimize performance include reviewing the:
- Review the product configuration
The first thing to do is review the product configuration. One simple way to do this is to look in your application control and server regions in SDSF. When each server starts, the runtime prints out the current configuration data in the job log.
- Internal tracing tips for the product
Product traces can be extremely helpful in detecting and diagnosing problems. By properly setting trace options, you can capture the information needed to detect problems without significant performance overhead.
- Verify that you are not collecting more diagnostic data than you need.
You should check your tracing options to verify that the ras_trace_defaultTracingLevel property is set to 0 or 1, and that the ras_trace_basic, and ras_trace_detail properties are not set.
How to view or set: In the administrative console:- Click .
- On the Configuration Tab check for any of these properties in the name field and observe their settings in the value field.
- To change the setting for one of these properties, click the property name the name field, and then specify the new setting in the value field. You can also describe the setting in the description field on this tab.
- To add one of these properties, click New, and then specify the property name the name field, and the property setting in the value field.
- If you use any level of tracing, including ras_trace_defaultTracingLevel=1, verify that the
ras_trace_outputLocation property is set to BUFFER.When the ras_trace_defaultTracingLevel property is set to 1 exceptions are written to the trace log as well as to the ERROR log.
- It is best to trace to CTRACE.
If you are tracing to SYSPRINT with ras_trace_defaultTracingLeve set to 3, you might experience an almost 100% throughput degradation. If you are tracing to CTRACE, however, you might only experience a 15% degradation in throughput.
- It is best to trace to CTRACE.
- Set the ras_trace_BufferCount property to 4 and the ras_trace_BufferSize
property to 128.
This setting reserves 512 KB of storage for the trace buffers, which is the minimum amount of storage that is allowed, and reduces memory requirements.
- Disable JRAS tracing. To disable JRAS tracing, look for the following lines in the trace.dat file that is pointed to by the JVM properties file:
Verify that both lines are set to disable, or delete the two lines.com.ibm.ejs.*=all=disable com.ibm.ws390.orb=all=disable
Note: If a value is specified for the ras_trace_outputLocation property, you might be tracing and not know it.
- Verify that you are not collecting more diagnostic data than you need.
- Location of executable programs tips
The next thing to review in the configuration is where your program code is located. IBM recommends that you install as much of the product code in LPA as is reasonable. This ensures that you have eliminated any unnecessary steplibs which can adversely affect performance. If you must use STEPLIBs, verify that any
STEPLIB DD
in the controller and servant procs do not point to any unnecessary libraries. Refer to UNIX System Services (USS) tuning tips for z/OS for USS shared file system tuning considerations.If you choose to not put most of the runtime in LPA, you might find that your processor storage gets a bigger workout as the load increases. At a minimum, the product starts three address spaces, so that any code that is not shared loads three copies rather than one. As the load increases, many more servants might start and contribute additional load on processor storage.
Review the
PATH
statement to ensure that only required programs are in the PATH and that the order of the PATH places frequently-referenced programs in the front. - Tuning for J2EE
applicationsSteps involved in tuning the J2EE applications performance include:
- Review your application design
You can track many performance problems back to the application design. Review the design to determine if it causes performance problems.