This section describes the relationship between the reloading functionality and the following functionality:
This subsection describes the relationship between the addition of the new requests to the pending list and the controlling of the concurrently executed threads in the Web application reloading functionality. For details on controlling the concurrently executed threads, see 2.15 Overview of controlling the concurrently executed threads in the uCosminexus Application Server Web Container Functionality Guide.
The new requests that are pending with the reloading functionality are controlled separately from the pending queue for controlling the concurrently executed threads for the Web applications or URL groups.
When the concurrently executed threads are controlled for Web containers, the processing threads of the new requests that can now be executed using the concurrently executed thread control for Web containers, are added to the pending list, after the reload processing starts.
Therefore, when the number of new requests pending with the reloading functionality reaches the maximum number of concurrently executed threads for the Web containers, the requests to the Web applications, other than the Web applications for which the reload processing was started, are also added to the pending list by the functionality for controlling the concurrently executed threads for Web containers.
When the processing of the requests being processed does not finish, you can specify the -t option with the cjreloadapp command to forcibly terminate the requests being processed and start the reload processing. However, the requests being processed can be forcibly terminated only by those methods in which the method timeout for monitoring the J2EE application execution time is applicable. Also, when method cancellation is set to operate after the method timeout occurs, the requests being processed are cancelled because of the timeout and the reloading process starts.
For details on the monitoring of the J2EE application execution time, see 5.3 Monitoring and cancelling the J2EE application execution time in the uCosminexus Application Server Operation, Monitoring, and Linkage Guide. For details on the settings for monitoring the J2EE application execution time, see 13.8.12 Settings for detecting updates and reloading the J2EE applications.
This point describes the reloading of the EJB applications.
If the interval for configuration file update lapses, the J2EE server waits for the processing of the requests being processed to finish and then starts the reload processing. If the processing of requests being processed is not over at this point, the method processes not terminated in a fixed time are reported as timed out during the monitoring of the J2EE application execution time, the method processing is cancelled, and then the reload processing can be started.
The following figure shows the process of reloading with the method cancellation functionality.
Figure 13-14 Reloading with the method cancellation functionality (For the EJB applications)
You obtain the time from the detection of updates in the configuration file to the start of the reload processing with the following formula:
Interval-for-configuration-file-update-(seconds) + Pending-time-of-requests-being-processed-(seconds)
This point describes the reloading of the Web applications.
If delayed reloading is specified and if the maximum specified delay time lapses, the J2EE server sets the new requests to a pending state, waits for the processing of the requests being processed to finish, and then starts the reload processing. If the processing of requests being processed is not over at this point, the method processes not terminated in a fixed time are reported as timed out during the monitoring of the J2EE application execution time, the method processing is cancelled, and then the reload processing can be started.
The following figure shows the process of reloading with the method cancellation functionality.
Figure 13-15 Reloading with the method cancellation functionality (For the Web applications)
You obtain the time from the detection of updates in the configuration file to the start of the reload processing with the following formula:
Interval-for-configuration-file-update-(seconds) + Maximum-delay-time-for-delayed-reloading-(seconds) + Pending-time-of-requests-being-processed-(seconds)