Hitachi

JP1 Version 12 JP1/Automatic Operation Configuration Guide 


4.8 Procedure to change the maximum number of plug-ins that can be executed concurrently

To change the maximum number of plug-ins that can be executed concurrently, you must restart the product.

You can change the maximum number of concurrently executable plug-ins that are included in normal tasks. The value that can be specified is 10, 50, or 100.

If you change the maximum number of plug-ins that can be executed concurrently, you must also review the amount of memory on the JP1/AO server.

To change the maximum number of plug-ins that can be executed concurrently:

  1. Estimate the maximum number of plug-ins that will be executed concurrently. Then, based on this estimation, secure the necessary amount of memory on the JP1/AO server. The following table lists the relationship between the number of concurrently executed plug-ins and the amount of memory required on the JP1/AO server.

    Table 4‒1: Relationship between the number of concurrently executed plug-ins and the amount of memory required on the JP1/AO server

    Number of concurrently executed plug-ins

    Amount of memory required on the JP1/AO server (GB)#

    Minimum

    Recommended

    10

    4

    6

    50

    5

    7

    100

    6

    8

    # If the upper limit values for the following types of processing might be reached at the same time, we recommend that you secure the recommended amount of memory:

    • User-response wait plug-ins (upper limit: 70)

    • Repeated execution plug-ins (upper limit: 20)

    • Debug tasks (upper limit: 10)

  2. For the plugin.threadPoolSize key in the user-specified properties file (config_user.properties), specify the number-of-concurrently-executed-plug-ins value (10, 50, or 100) in the following format:

    plugin.threadPoolSize = number-of-concurrently-executed-plug-ins

  3. Stop the JP1/AO services.

    For non-cluster systems:

    Execute the hcmds64srv command with the stop option specified.

    For cluster systems:

    Use the cluster software to bring the services offline.

  4. Start the JP1/AO services.

    For non-cluster systems:

    Execute the hcmds64srv command with the start option specified.

    For cluster systems:

    Use the cluster software to bring the services online.

Tip

When plug-ins requiring a long execution time are executed concurrently, if the maximum number of concurrently executable plug-ins is reached, other plug-ins are placed in a wait state and cannot start even if the system has sufficient processing performance. In such a case, you can reduce unnecessary waits to shorten the overall processing time by increasing the maximum number of concurrently executable plug-ins. Note that you cannot shorten the processing time by this method in cases requiring all JP1/AO processing ability, such as when many plug-ins requiring a short execution time are executed.