2.5.6 Considering reduction of delays in job status transitions
When the manager host communicates with multiple execution agents, if a communication failure occurs during the response to one or more execution agents, the manager host continuously responds to the execution agents for a maximum of 10 minutes. At this time, job execution might be delayed. For example, the job status might not change for 10 minutes. Due to the impact of the communication failure, a delay of 10 or more minutes might also occur during the status transition of a job executed on an execution agent that is available for normal communication.
In this case, you can suppress response from the manager host to execution agents that cannot respond normally so that the manager host responds to notifications only from normal execution agents. This can reduce delays in job status transitions. This function is called function for reducing job status transition delays.
- Organization of this subsection
(1) Overview of the function for reducing job status transition delays
The function for reducing job status transition delays can suppress response to an execution agent on which a communication failure has occurred. This can reduce delays in the status transition of jobs executed on normal execution agents. The following figure shows an example of job status transition when a failure occurs during communication with execution agents.
|
This function applies to communication shown in the following table.
Service name |
Port number |
Purpose |
---|---|---|
jp1ajs2qman |
20241/tcp |
To receive job#1 registrations To receive a job's#1 standard output file and standard error output file To accept requests for execution of job execution commands#2 other than jp1exec and jp1exit To link with JP1/OJE for VOS3 (to receive job registrations) |
jp1ajs2qnfy |
20243/tcp |
To receive job#1 start/end notifications from JP1/AJS3 - Agent |
jp1ajs2eamgr |
20246/tcp |
JP1/AJS3 - Manager event job execution |
- #1
-
Standard jobs (other than queueless jobs), HTTP connection jobs, action jobs, and custom jobs.
- #2
-
See the table of job execution commands in 1.5 Commands in the manual JP1/Automatic Job Management System 3 Command Reference.
The function for reducing job status transition delays is disabled by default. To enable the function, set the following two environment setting parameters:
-
ResponseTimeout
-
ReduceStateTransitionDelay
For details about how to enable the function for reducing job status transition delays, see 20.8.2 Details of the environment setting parameters for communication control in the JP1/Automatic Job Management System 3 Configuration Guide.
(2) Jobs subject to function for reducing job status transition delays
The function for reducing job status transition delays can be applied to the following types of jobs:
-
PC job (with the exception of queueless jobs)
-
Unix job (with the exception of queueless jobs)
-
Action job (with the exception of queueless jobs)
-
Submit Job
-
Queue Job
-
Custom job
-
HTTP connection job
-
Passing information setting job
-
Flexible Job
-
Event Job
(3) Interval for monitoring communication with execution agents
If the function for reducing job status transition delays is enabled and a communication failure is detected on an execution agent, the manager host suppresses a response to that execution agent for 5 minutes after the failure is detected. If the communication recovers after 5 minutes have passed, the manager host terminates the suppression of response. If the communication has not recovered, the manager host suppresses response for a further 5 minutes. The following shows the interval for monitoring communication with execution agents.
|
-
The amount of time that can elapse before a communication failure is detected when the manager host responds to a notification from an execution agent. You can change this value in the ResponseTimeout environment setting parameter. The default is 10 seconds.
-
The amount of time for which the manager host suppresses response to the execution agent. This value is fixed to 5 minutes.
-
The interval at which the execution agent resends a notification to the manager host. You can specify this value for each job by using an environment setting parameter. The following table shows the applicable jobs and environment setting parameters.
Table 2‒45: Applicable jobs and environment setting parameters Job
Environment setting parameter
Standard job (with the exception of queueless jobs)
-
NotfyJobStateInterval
-
NotfyJobStateCount
HTTP connection job
Action job
Custom job
Event Job
-
NotificationRetryInterval
-
NotificationRetryCount
-
If a response to the execution agent fails and a communication failure is detected, the manager host outputs the KNAC0403-W message to the integrated trace log. The KNAC0403-W message is output for each IP address of the execution agent where a communication failure occurred. If communication with the execution agent recovers, the KNAC0402-I message is output to the integrated trace log. If a communication failure occurs again after communication recovers or after the JP1/AJS3 service stops, the KNAC0403-W message is output again.
(4) Note on the function for reducing job status transition delays
Using the function for reducing job status transition delays might cause a delay in detecting a recovery from communication failure. The following figure shows when a recovery from a communication failure can be detected depending on whether the function for reducing job status transition delays is disabled or enabled.
|