Problem | If the system is meant to be under load, there may be a bottleneck within the application itself ( such as over synchronization ), or elsewhere in the system. If this problem occurs in a test environment, the load driver may have reached its maximum limit. |
User response | Consult the WebSphere Application Server InfoCenter for information on deadlocks and application performance tuning. |
Problem | Caching all prepared statements typically improves performance. In general, the size of the cache should equal the number of prepared statements used in the application. |
User response | From the administrative console, click: JDBC Providers > JDBC_provider > Data Sources > data_source and increase the Statement Cache Size field. For more information about tuning datasource connection pools open the following url: http://www14.software.ibm.com/webapp/wsbroker/redirect?version=pix&product=was-base-dist&topic=datobjtune |
Problem | When a growable thread pool expands beyond the maximum size, those additional threads will not be reused and will be discarded from the pool. Increasing the thread pool maximum size to the average pool size can improve performance significantly. If this message appears repeatedly, look for continued growth of the thread pool, which often indicates a leak or deadlock. |
User response | For the Web container, click: Servers > Application Servers > Server > Thread Pools. For the ORB Service, click: Servers > Application Servers > Server > Container Services > ORB Service > Thread Pool. For more information on tuning your Application Server open the following url: http://www14.software.ibm.com/webapp/wsbroker/redirect?version=pix&product=was-base-dist&topic=tprf_tuneappserv. For more information about queueing in WebSphere open the following url: http://www14.software.ibm.com/webapp/wsbroker/redirect?version=pix&product=was-base-dist&topic=rprf_queuetip |
Problem | If this message appears repeatedly, look for continued growth of the thread pool, which often indicates a leak or deadlock. This message may also indicate a temporary load surge. However, if the thread pool remains steadily at this size, then the large number of threads may hurt performance. Consider testing a bounded thread pool of a smaller size. |
User response | For the Web container, click: Servers > Application Servers > Server > Thread Pools. For the ORB Service, click: Servers > Application Servers > Server > Container Services > ORB Service > Thread Pool. Consult the WebSphere Application Server InfoCenter for information on deadlocks. For more information on tuning your Application Server open the following url: http://www14.software.ibm.com/webapp/wsbroker/redirect?version=pix&product=was-base-dist&topic=tprf_tuneappserv. For more information about queueing in WebSphere open the following url: http://www14.software.ibm.com/webapp/wsbroker/redirect?version=pix&product=was-base-dist&topic=rprf_queuetip |
Problem | The CPU utilization of the system is high. Decreasing the size of the thread pool will free resources and may improve performance. |
User response | For the Web container, click: Servers > Application Servers > Server > Thread Pools. For the ORB Service, click: Servers > Application Servers > Server > Container Services > ORB Service > Thread Pool. For more information on tuning your Application Server open the following url: http://www14.software.ibm.com/webapp/wsbroker/redirect?version=pix&product=was-base-dist&topic=tprf_tuneappserv. For more information about queueing in WebSphere open the following url: http://www14.software.ibm.com/webapp/wsbroker/redirect?version=pix&product=was-base-dist&topic=rprf_queuetip |
Problem | The utilization of the thread pool is high, yet the CPU utilization is low. It is possible that the thread pool size is too small, and might be unnecessarily limiting the performance of your system. WARNING: Increasing the size of the pool can also hurt performance. Test carefully. Optimal performance is usually obtained when the thread pool size is just large enough to support full CPU utilization. It is normal to see high utilization of the thread pool. |
User response | For the Web container, click: Servers > Application Servers > Server > Thread Pools. For the ORB Service, click: Servers > Application Servers > Server > Container Services > ORB Service > Thread Pool. For more information on tuning your Application Server open the following url: http://www14.software.ibm.com/webapp/wsbroker/redirect?version=pix&product=was-base-dist&topic=tprf_tuneappserv. For more information about queueing in WebSphere open the following url: http://www14.software.ibm.com/webapp/wsbroker/redirect?version=pix&product=was-base-dist&topic=rprf_queuetip |
Problem | Decreasing the size supports better pooling and frees memory resources. |
User response | From the administrative console, click: Resources > JDBC Providers > JDBC_provider > Data Sources > data_source > Connection pool properties. For more information about tuning datasource connection pools open the following url: http://www14.software.ibm.com/webapp/wsbroker/redirect?version=pix&product=was-base-dist&topic=datobjtune |
Problem | It is possible that the connection pool is unnecessarily limiting the performance of your system. WARNING: Increasing the size of the pool can also hurt performance. Test carefully. Optimal performance is usually obtained when the connection pool is just large enough. In general, expect to see high utilization of the thread pool. |
User response | From the administrative console, click: Resources > JDBC Providers > JDBC_provider > Data Sources > data_source > Connection pool properties. For more information about tuning datasource connection pools open the following url: http://www14.software.ibm.com/webapp/wsbroker/redirect?version=pix&product=was-base-dist&topic=datobjtune |
Problem | The system appears to be under load (CPU is very busy,) yet the number of threads used within the pool is small. Decreasing the size of the thread pool will free resources and may improve performance. |
User response | From the administrative console, click: Resources > JDBC Providers > JDBC_provider > Data Sources > data_source > Connection pool properties. For more information about tuning datasource connection pools open the following url: http://www14.software.ibm.com/webapp/wsbroker/redirect?version=pix&product=was-base-dist&topic=datobjtune |
Problem | There might be a problem with your session repository, your sessions might be too large, or you might be writing to the repository more often than necessary. If you use multi-row sessions, the times above are per attribute, rather than per session. |
User response | Consult the application provider about reducing the session size. For more information about Application Design open the following URL : http://www14.software.ibm.com/webapp/wsbroker/redirect?version=pix&product=was-base-dist&topic=resourcesforlearning. For more information about best practices for using HTTP Sessions open the following url: http://www14.software.ibm.com/webapp/wsbroker/redirect?version=pix&product=was-base-dist&topic=rpersess. For more information on Session Management Tuning open the following url: http://www14.software.ibm.com/webapp/wsbroker/redirect?version=pix&product=was-base-dist&topic=rpersest |
Problem | Large sessions can hurt performance. The major costs are serialization, communication with the repository, and the amount of memory used. Consider revising your application to reduce the session size. |
User response | Consult the application provider about reducing the session size. For more information about Application Design open the following URL : http://www14.software.ibm.com/webapp/wsbroker/redirect?version=pix&product=was-base-dist&topic=resourcesforlearning. For more information about best practices for using HTTP Sessions open the following url: http://www14.software.ibm.com/webapp/wsbroker/redirect?version=pix&product=was-base-dist&topic=rpersess |
Problem | There might be a problem with your session repository, or your sessions might be too large. If you use multi-row sessions, the times above are per attribute, rather than per session. |
User response | Consult the application provider about reducing the session size. For more information about Application Design open the following URL : http://www14.software.ibm.com/webapp/wsbroker/redirect?version=pix&product=was-base-dist&topic=resourcesforlearning. For more information about best practices for using HTTP Sessions open the following url: http://www14.software.ibm.com/webapp/wsbroker/redirect?version=pix&product=was-base-dist&topic=rpersess. For more information on Session Management Tuning open the following url: http://www14.software.ibm.com/webapp/wsbroker/redirect?version=pix&product=was-base-dist&topic=rpersest |
Problem | Large sessions can hurt performance. The major costs are serialization, communication with the repository, and the amount of memory used. Consider revising your application to reduce the session size. |
User response | Consult the application provider about reducing the session size. For more information about Application Design open the following URL : http://www14.software.ibm.com/webapp/wsbroker/redirect?version=pix&product=was-base-dist&topic=resourcesforlearning. For more information about best practices for using HTTP Sessions open the following url: http://www14.software.ibm.com/webapp/wsbroker/redirect?version=pix&product=was-base-dist&topic=rpersess |
Problem | In this situation, the overflow session cache is used instead of the main session cache, possibly hurting performance. Check that the session growth is bounded. In general, the average number of live sessions is approximately the rate of session creation times the average lifetime of a session. |
User response | From the administrative console, click: Servers > Application Servers > Server > Web Container Settings > Web Container > Session Management. For more information on Session Management Tuning open the following url: http://www14.software.ibm.com/webapp/wsbroker/redirect?version=pix&product=was-base-dist&topic=rpersest |
Problem | When the session cache is full and can not overflow, getSession will return an invalid session. The average number of live sessions is approximately the rate of session creation times the average lifetime of a session. If this situation was caused by a temporary load surge, consider allowing the cache to overflow. Note that if an application has unbounded session growth, allowing overflow will result in a memory leak. |
User response | From the administrative console, click: Servers > Application Servers > Server > Web Container Settings > Web Container > Session Management. For more information on Session Management Tuning open the following url: http://www14.software.ibm.com/webapp/wsbroker/redirect?version=pix&product=was-base-dist&topic=rpersest |
Problem | An alert was issued recommending a setting increase. However there is not enough memory available in the JVM heap to implement the change. Consider changing the JVM heap size, the time spent in garbage collection and the memory usage. |
User response | From the administrative console, click: Application Servers > Server > Java and Process Management > Process Definition > Java Virtual Machine. |
Problem | Data access is not available for this data type. |
User response | See the Troubleshooting section of the InfoCenter. |
Problem | When the dynamic cache is not large enough, the server must recalculate the evicted entries. However, in many systems, there is not enough memory in the heap for a cache large enough to hold all cacheable entries. Use the Tivoli Performance Viewer or other monitoring software to watch heap utilization as you adjust the dynamic cache size. |
User response | From the administrative console, click: Servers > Application Servers > Server > Container Services > Dynamic Cache Service. |
Problem | The dynamic cache is not using all the available entries. If the server is operating under normal load, consider reducing the number of dynamic cache entries so that these resources can be allocated elsewhere. |
User response | From the administrative console, click: Servers > Application Servers > Server > Container Services > Dynamic Cache Service. |
Problem | There are many reasons why an application may spend too much time in garbage collection. Often, it indicates an application with many short-lived objects. However, it can also indicate a heap that is too small, causing the JVM's memory manager to thrash. Increasing the heap size will help in this case, by lengthening the time between garbage collection calls, but will also increase the time needed for each garbage collection. |
User response | From the administrative console, click: Application Servers > Server > Java and Process Management > Process Definition > Java Virtual Machine. For more information on tuning Java Virtual Machines open the following url: http://www14.software.ibm.com/webapp/wsbroker/redirect?version=pix&product=was-base-dist&topic=tprf_tunejvm_v61. For more information Java memory tuning tips open the following url: http://www14.software.ibm.com/webapp/wsbroker/redirect?version=pix&product=was-base-dist&topic=rprf_javamemory |
Problem | Session size contributes to application performance. If the data is available, the performance advisor can give advice on the average session size. |
User response | Enable the PMI counter Session Object Size if not using distributed sessions. |
Problem | Large sessions can hurt performance. The major costs are serialization, and the amount of memory used. Consider revising your application to reduce the average session size. |
User response | Contact the application developer about decreasing the average session size. For more information about Application Design open the following URL : http://www14.software.ibm.com/webapp/wsbroker/redirect?version=pix&product=was-base-dist&topic=resourcesforlearning. For more information about best practices for using HTTP Sessions open the following url: http://www14.software.ibm.com/webapp/wsbroker/redirect?version=pix&product=was-base-dist&topic=rpersess |
Problem | Large sessions can hurt performance. The major costs are serialization, and the amount of memory used. Consider revising your application to reduce the maximum session size. |
User response | Contact the application developer about decreasing the maximum session size. For more information about Application Design open the following URL : http://www14.software.ibm.com/webapp/wsbroker/redirect?version=pix&product=was-base-dist&topic=resourcesforlearning. For more information about best practices for using HTTP Sessions open the following url: http://www14.software.ibm.com/webapp/wsbroker/redirect?version=pix&product=was-base-dist&topic=rpersess |
Problem | The performance advisor does not need the "PMI monitored statistic set" to be "all". Lowering the "PMI monitored statistic set" to "extended" can provide additional performance boost. |
User response | If other applications are not using this PMI data, then consider lowering the "PMI monitored statistic set" to "extended". For more information about PMI open the following url: http://www14.software.ibm.com/webapp/wsbroker/redirect?version=pix&product=was-base-dist&topic=tprf_pmi_encoll |
Problem | The connection pool size is fine and the pool usage is good. |
User response | No action is required. |
Problem | The current algorithm only decreases the Minimum/Maximum Pool Size settings if the current pool size is equal to the MinPoolSize setting. |
User response | No action is required. |
Problem | The CPU utilization of this system is unusually low. Please run the performance advisor with a representative workload. If CPU utilization is expected to be this low, then consider decreasing the size of the thread pool. If not, the bottleneck may exist elsewhere in the system, preventing work from reaching the WebSphere Application Server. |
User response | To decrease the size of the thread pool: For the Web Container, click: Servers > Application Servers > Server > Thread Pools. For the ORB Service, click: Servers > Application Servers > Server > Container Services > ORB Service > Thread Pool. |
Problem | Connection pool size looks good. |
User response | No action is required. |
Problem | The CPU is already heavily utilized. Increasing the number of threads in the pool increases the time spent switching between threads. |
User response | No action is required. |
Problem | The current thread pool size is less than the maximum defined size. |
User response | No action is required. |
Problem | The session cache size is well configured. |
User response | No action is required. |
Problem | The session cache has enough entries to store the average number of live sessions. |
User response | No action is required. |
Problem | The session cache size is small enough that resources are not wasted. |
User response | No action is required. |
Problem | The size of the session objects written to the persistent store is within reasonable limits. |
User response | No action is required. |
Problem | The size of the session objects read from the persistent store is within reasonable limits. |
User response | No action is required. |
Problem | Reading large sessions can cause serious performance problems. |
User response | No action is required. |
Problem | Writing large sessions can cause serious performance problems. |
User response | No action is required. |
Problem | Session cache size looks good. |
User response | No action is required. |
Problem | The size of the prepared statement cache is set appropriately. |
User response | No action is required. |
Problem | Either the thread pool size is negative, or the number of active threads is greater than the thread pool size. |
User response | Check your thread pool settings in the administrative console: For the Web container, click: Servers > Application Servers > Server > Thread Pools. For the ORB Service, click: Servers > Application Servers > Server > Container Services > ORB Service > Thread Pool |
Problem | Remember that the performance advisor limits the number of times a particular warning is issued. A tuning problem can still exist, even though no more messages are issued. |
User response | No action is required. |
Problem | The Runtime Performance Advisor enables this data point for future use. |
User response | From the administrative console, click: Application Servers > Server > Performance Monitoring Infrastructure (PMI). Note: Setting the JVM module to all requires enabling the Java Virtual Machine Profiling Interface (JVMPI). See the InfoCenter for more on this topic. |
Problem | To fully utilize the performance advisor, enable monitoring for these modules. |
User response | From the administrative console, click: Application Servers > Server > Performance Monitoring Infrastructure (PMI). Note: Setting the JVM module to MAX requires enabling the Java Virtual Machine Profiling Interface (JVMPI). See the InfoCenter for more on this topic. |
Problem | The Performance Monitoring Infrastructure (PMI) is not enabled for this server. |
User response | From the administrative console, click: Application Servers > Server > Performance Monitoring Infrastructure (PMI). Enable the Performance Monitoring Infrastructure (PMI) and restart the server. |
Problem | The Runtime Performance Advisor, advice, or Performance Monitoring Infrastructure (PMI) data point might have just been enabled. |
User response | No action is required. |
Problem | To use the Runtime Performance Advisor, restart the application server. |
User response | From the WebSphere Application Server installation directory, go to the bin directory and type stopserver.bat to stop the server and then type startserver.bat to start the server. |
Problem | A general error was detected. |
User response | Locate the logs for WebSphere Application Server, in the logs directory, under the installation root directory. Each server instance has its own directory in the logs directory. Enable tracing in the administrative console, by clicking: Application Servers > Server > Logging and Tracing > Diagnostic Trace Service. |
Problem | An unknown error was detected. |
User response | Locate the logs for WebSphere Application Server, in the logs directory, under the installation root directory. Each server instance has its own directory in the logs directory. Enable tracing in the administrative console, by clicking: Application Servers > Server > Logging and Tracing > Diagnostic Trace Service. |
Problem | There are no problems with the number of dynamic cache entries allocated. |
User response | No action is required. |
Problem | The Java virtual machine is spending a reasonable amount of time garbage collecting. |
User response | No action is required. |
Problem | The in-memory session sizes are all reasonable. |
User response | No action is required. |
Problem | The Performance Monitoring Infrastructure (PMI) is not enabled for this nodeagent. |
User response | From the administrative console, click: System Administration > Node Agents > nodeagent > Performance Monitoring Infrastructure (PMI). Enable the Performance Monitoring Infrastructure (PMI) and restart the nodeagent. |
Problem | Performance Runtime Advisor expected to get an admin client and did not. |
User response | Check that Node Agent is running and that process discovery has occurred on the application server. |
Problem | The Runtime Performance Advisor registers for notifications of runtime changes. For some reason it was unable to do so. |
User response | If running in ND please make sure that the Deployment Manager is still running |
Problem | Internal WebSphere error |
User response | Contact support |
Problem | Internal WebSphere error |
User response | Contact support |
Problem | The Runtime Performance Advisor registers for notifications of runtime changes. For some reason it was unable to do so. |
User response | If running in ND please make sure that the Deployment Manager is still running |
Problem | The Runtime Peformance Advisor uses the runtime settings in order to help determine performance problems. This data was not available so that persistent configuration will be used instead of the runtime configuration. |
User response | None |
Problem | There was a problem getting the Application Server Configuration details from the Repository. |
User response | Please contact IBM support. |
Problem | There was a problem getting the Application Server Session Configuration details from the Repository. |
User response | Please contact IBM support. |
Problem | There was a problem getting the Resource from the config root. |
User response | Please contact IBM support. |
Problem | A general error was detected. |
User response | Locate the logs for WebSphere Application Server, in the logs directory, under the installation root directory. Each server instance has its own directory in the logs directory. Enable com.ibm.ws.performance.tuning.serverAlert.*=all=enabled tracing in the administrative console, by clicking: Application Servers > Server > Logging and Tracing > Diagnostic Trace Service. |
Problem | An error occurred while running performance advisor. |
User response | Restart performance advisor. |
Problem | If the initial JVM heap size is too small, the heap has to expand more often than desired, reducing performance when the application server is first initialized. |
User response | From the administrative console, click: Application Servers > Server > Java and Process Management > Process Definition > Java Virtual Machine. For more information on tuning Java Virtual Machines open the following url: http://www14.software.ibm.com/webapp/wsbroker/redirect?version=pix&product=was-base-dist&topic=tprf_tunejvm_v61. For more information Java memory tuning tips open the following url: http://www14.software.ibm.com/webapp/wsbroker/redirect?version=pix&product=was-base-dist&topic=rprf_javamemory |
Problem | If the minimum heap size is too large, the JVM cannot reduce the heap to an appropriate size. Garbage collection takes more time, causing longer interruptions in application processing. |
User response | From the administrative console, click: Application Servers > Server > Java and Process Management > Process Definition > Java Virtual Machine. For more information on tuning Java Virtual Machines open the following url: http://www14.software.ibm.com/webapp/wsbroker/redirect?version=pix&product=was-base-dist&topic=tprf_tunejvm_v61. For more information Java memory tuning tips open the following url: http://www14.software.ibm.com/webapp/wsbroker/redirect?version=pix&product=was-base-dist&topic=rprf_javamemory |
Problem | If the maximum heap size is too small, the JVM does not have enough room to manage the heap efficiently. Performance degrades, and the application can fail. |
User response | From the administrative console, click: Application Servers > Server > Java and Process Management > Process Definition > Java Virtual Machine. For more information on tuning Java Virtual Machines open the following url: http://www14.software.ibm.com/webapp/wsbroker/redirect?version=pix&product=was-base-dist&topic=tprf_tunejvm_v61. For more information Java memory tuning tips open the following url: http://www14.software.ibm.com/webapp/wsbroker/redirect?version=pix&product=was-base-dist&topic=rprf_javamemory |
Problem | If the heap size is too large for the amount of physical memory on your machine, then your system will page, reducing performance. Monitor your system. If this situation occurs, attempt to reduce your application footprint or add physical memory to your system. |
User response | From the administrative console, click: Application Servers > Server > Java and Process Management > Process Definition > Java Virtual Machine. For more information on tuning Java Virtual Machines open the following url: http://www14.software.ibm.com/webapp/wsbroker/redirect?version=pix&product=was-base-dist&topic=tprf_tunejvm_v61. For more information Java memory tuning tips open the following url: http://www14.software.ibm.com/webapp/wsbroker/redirect?version=pix&product=was-base-dist&topic=rprf_javamemory |
Problem | The JVM heap size settings specify an initial heap size that is larger than the maximum heap size. |
User response | From the administrative console, click: Application Servers > Server > Java and Process Management > Process Definition > Java Virtual Machine. For more information on tuning Java Virtual Machines open the following url: http://www14.software.ibm.com/webapp/wsbroker/redirect?version=pix&product=was-base-dist&topic=tprf_tunejvm_v61. For more information Java memory tuning tips open the following url: http://www14.software.ibm.com/webapp/wsbroker/redirect?version=pix&product=was-base-dist&topic=rprf_javamemory |
Problem | Having the JIT disabled will reduce performance of the Application Server. |
User response | From the administrative console, click: Application Servers > Server > Java and Process Management > Process Definition > Java Virtual Machine. For more information on tuning Java Virtual Machines open the following url: http://www14.software.ibm.com/webapp/wsbroker/redirect?version=pix&product=was-base-dist&topic=tprf_tunejvm_v61. |
Problem | Having the JIT enabled improves performance. |
User response | No action is required |
Problem | Enabled trace specifications may significantly reduce the performance of the Application Server. |
User response | From the administrative console, click: Application Servers > Server > Change Log or Trace Levels and disable all unwanted trace specifications |
Problem | The value specified for the minimum thread pool of the Web container is greater than the value for the maximum thread pool. |
User response | From the administrative console, click: Servers > Application Servers > Server > Thread Pools. For more information on tuning your Application Server open the following url: http://www14.software.ibm.com/webapp/wsbroker/redirect?version=pix&product=was-base-dist&topic=tprf_tuneappserv. For more information about queueing in WebSphere open the following url: http://www14.software.ibm.com/webapp/wsbroker/redirect?version=pix&product=was-base-dist&topic=rprf_queuetip |
Problem | In general, a very large thread pool hurts performance, although it might be necessary for some applications. See the WebSphere Application Server Tuning Guide for more information. |
User response | From the administrative console, click: Servers > Application Servers > Server > Thread Pools. For more information on tuning your Application Server open the following url: http://www14.software.ibm.com/webapp/wsbroker/redirect?version=pix&product=was-base-dist&topic=tprf_tuneappserv. For more information about queueing in WebSphere open the following url: http://www14.software.ibm.com/webapp/wsbroker/redirect?version=pix&product=was-base-dist&topic=rprf_queuetip |
Problem | In general, a very large thread pool hurts performance, although it might be necessary for some applications. See the WebSphere Application Server Tuning Guide for more information. |
User response | From the administrative console, click: Servers > Application Servers > Server > Thread Pools. For more information on tuning your Application Server open the following url: http://www14.software.ibm.com/webapp/wsbroker/redirect?version=pix&product=was-base-dist&topic=tprf_tuneappserv. For more information about queueing in WebSphere open the following url: http://www14.software.ibm.com/webapp/wsbroker/redirect?version=pix&product=was-base-dist&topic=rprf_queuetip |
Problem | The value specified for the minimum thread pool of the ORB is greater than the value for the maximum thread pool. |
User response | From the administrative console, click: Servers > Application Servers > Server > Container Services > ORB Service > Thread Pool. For more information on tuning your Application Server open the following url: http://www14.software.ibm.com/webapp/wsbroker/redirect?version=pix&product=was-base-dist&topic=tprf_tuneappserv. For more information about queueing in WebSphere open the following url: http://www14.software.ibm.com/webapp/wsbroker/redirect?version=pix&product=was-base-dist&topic=rprf_queuetip |
Problem | In general, a very large thread pool hurts performance, although it might be necessary for some applications. See the WebSphere Application Server Tuning Guide for more information. |
User response | From the administrative console, click: Servers > Application Servers > Server > Container Services > ORB Service > Thread Pool. For more information on tuning your Application Server open the following url: http://www14.software.ibm.com/webapp/wsbroker/redirect?version=pix&product=was-base-dist&topic=tprf_tuneappserv. For more information about queueing in WebSphere open the following url: http://www14.software.ibm.com/webapp/wsbroker/redirect?version=pix&product=was-base-dist&topic=rprf_queuetip |
Problem | In general, a very large thread pool hurts performance, although it might be necessary for some applications. See the WebSphere Application Server Tuning Guide for more information. |
User response | From the administrative console, click: Servers > Application Servers > Server > Container Services > ORB Service > Thread Pool. For more information on tuning your Application Server open the following url: http://www14.software.ibm.com/webapp/wsbroker/redirect?version=pix&product=was-base-dist&topic=tprf_tuneappserv. For more information about queueing in WebSphere open the following url: http://www14.software.ibm.com/webapp/wsbroker/redirect?version=pix&product=was-base-dist&topic=rprf_queuetip |
Problem | The value specified for the minimum connection pool of the data source is greater than the maximum pool size. |
User response | From the administrative console, click: JDBC Providers > JDBC_provider > Data Sources > data_source > Connection pool properties. For more information about queueing in WebSphere open the following url: http://www14.software.ibm.com/webapp/wsbroker/redirect?version=pix&product=was-base-dist&topic=rprf_queuetip. For more information about tuning datasource connection pools open the following url: http://www14.software.ibm.com/webapp/wsbroker/redirect?version=pix&product=was-base-dist&topic=datobjtune |
Problem | In general, a very large connection pool hurts performance, although it might be necessary for some applications. See the WebSphere Application Server Tuning Guide for more information. |
User response | From the administrative console, click: JDBC Providers > JDBC_provider > Data Sources > data_source > Connection pool properties. For more information about queueing in WebSphere open the following url: http://www14.software.ibm.com/webapp/wsbroker/redirect?version=pix&product=was-base-dist&topic=rprf_queuetip. For more information about tuning datasource connection pools open the following url: http://www14.software.ibm.com/webapp/wsbroker/redirect?version=pix&product=was-base-dist&topic=datobjtune |
Problem | In general, a very large connection pool hurts performance, although it might be necessary for some applications. See the WebSphere Tuning Guide for more information. |
User response | From the administrative console, click: JDBC Providers > JDBC_provider > Data Sources > data_source > Connection pool properties. For more information about queueing in WebSphere open the following url: http://www14.software.ibm.com/webapp/wsbroker/redirect?version=pix&product=was-base-dist&topic=rprf_queuetip. For more information about tuning datasource connection pools open the following url: http://www14.software.ibm.com/webapp/wsbroker/redirect?version=pix&product=was-base-dist&topic=datobjtune |
Problem | A large prepared statement cache is usually a problem only if memory is constrained and there are many infrequently used prepared statements. In this case, the statements are not discarded, but consume memory, which can be used more effectively. In general, the size of the cache equals the number of prepared statements used in the application. |
User response | From the administrative console, click: JDBC Providers > JDBC_provider > Data Sources > data_source. For more information about queueing in WebSphere open the following url: http://www14.software.ibm.com/webapp/wsbroker/redirect?version=pix&product=was-base-dist&topic=rprf_queuetip. For more information about tuning datasource connection pools open the following url: http://www14.software.ibm.com/webapp/wsbroker/redirect?version=pix&product=was-base-dist&topic=datobjtune |
Problem | The DB2 Performance Configuration Wizard can calculate and suggest configuration parameters to optimize the performance of your DB2 database. |
User response | To access the DB2 Performance Configuration Wizard, right click on the database object in the DB2 Control Center and select Configure Performance Using Wizard. |
Problem | See the WebSphere Application Server InfoCenter for more information on using dynamic caching. |
User response | From the administrative console, click: Servers > Application Servers > Server > Container Services > Dynamic Cache Service. |
Problem | See the WebSphere Application Server InfoCenter for more information on servlet caching. |
User response | From the administrative console, click: Servers > Application Servers > Server > Web Container Settings > Web Container. For more information about Servlet Caching open the following url: http://www14.software.ibm.com/webapp/wsbroker/redirect?version=pix&product=was-base-dist&topic=cdyn_servletcaching |
Problem | An error occurred reading the configuration through the configuration service, or configuration data is not in the log file, if in log replay mode. |
User response | none |
Problem | Over a period of time the amount of free memory appears to be decreasing, or there is consistently insufficent free memory in the heap, indicating that you may have a memory leak. |
User response | Use tooling to further analyze your memory useage over time, for more information open the following link for Diagnosing out-of-memory errors and Java heap memory leaks: http://www14.software.ibm.com/webapp/wsbroker/redirect?version=pix&product=was-base-dist&topic=ttrb_mdd4j |
Problem | Over a period of time the amount of free memory appears to be decreasing, or there is consistently insufficent free memory in the heap, indicating that you may have a memory leak. |
User response | Use tooling to further analyze your memory useage over time, for more information open the following link for Diagnosing out-of-memory errors and Java heap memory leaks: http://www14.software.ibm.com/webapp/wsbroker/redirect?version=pix&product=was-base-dist&topic=ttrb_mdd4j |
Problem | There did not appear to be a GC cycle. The application does not appear to be in use. |
User response | No action is required |
Problem | Memory usage patterns do not indicate a memory leak |
User response | No action is required |
Problem | Memory usage patterns have indicated a potential problem |
User response | No action is required |
Problem | In order to provide early notification of a suspicous memory activity short and long data collection periods are used. Sometimes a short period issues a notification which is later determined to be false when looking at the larger amount of data contained within a long period. |
User response | No action is required |
Problem | It is possible that a memory leak is causing the size of the Java process to grow in an unbounded manner. |
User response | Either consider adjusting the memory pool size, or use tooling to further analyze your memory useage over time, for more information open the following link for Diagnosing out-of-memory errors and Java heap memory leaks: http://www14.software.ibm.com/webapp/wsbroker/redirect?version=pix&product=was-base-dist&topic=ttrb_mdd4j |
Problem | Memory usage patterns do not indicate a memory leak |
User response | No action is required |