2.1.2 Troubleshooting problems related to service startup
- Organization of this subsection
(1) A JP1/AJS3 service has not started
Possible causes are as follows:
-
If the KAVU5285-E message (There is no the database table, or it is short of the system resources. (reason-location)) is output to the integrated trace log:
If you are using QUEUE jobs or submit jobs, the job execution environment database for QUEUE jobs and submit jobs might not have been created correctly. Use the jpqimport command to create or re-create the job execution environment database for QUEUE jobs and submit jobs. For details about how to create or re-create the database, see 2.12.2 Procedure for re-creating the execution environment database for QUEUE jobs and submit jobs.
-
If the KAVU5284-E message (It is short of the system resources. (reason-location)) is output to the integrated trace log:
System resources, such as semaphores, required for JP1/AJS3 operation might not be sufficient.
Check the estimate for system resources, make sure that system resources are sufficient, and then restart JP1/AJS3.
-
If you start a JP1/AJS3 service when memory is insufficient, the KAVU1203-E message (The agent process could not be started. (Reason code: 12)) or the KAVU1204-E message (The manager process could not be started. (Reason code: 12)) might be output to the integrated trace log. If either message is output, reconsider the memory estimate. If any unnecessary applications are running, stop them and restart the JP1/AJS3 service.
-
When you start a JP1/AJS3 service, the KAVU1203-E message (The agent process could not be started. (Reason code: 0xffffffff)) or the KAVU1204-E message (The manager process could not be started. (Reason code: 0xffffffff)) might be output to the integrated trace log. If either message is output, initialization of the JP1/AJS3 service might have failed. Check the message that is output immediately before this message in the integrated trace log, eliminate the cause of the error, and then restart the JP1/AJS3 service.
-
When the JP1/AJS3 service is started, the following messages might be output to the integrated trace log:
-
KAVU1103-I Process monitor (logical-host-name) is already running on the same host.
-
KAVU4111-E Job queuing control (logical-host-name) or jpqimport command is already running on the same host.
-
KAVS0500-E Scheduler service has already started.
If any of the messages above are output, the JP1/AJS3 service might have terminated abnormally without stopping JP1/AJS3 processes. In this case, forcibly terminate the remaining JP1/AJS3 processes, and then restart the JP1/AJS3 service as described below.
- In Windows:
-
On both the physical and logical hosts, stop the JP1/AJS3 service, and then use the task manager to check whether JP1/AJS3 processes remain. If JP1/AJS3 processes remain, use the task manager to terminate them forcibly or restart the system.
- In UNIX:
-
If the JP1/AJS3 service on the physical host cannot be started, stop the JP1/AJS3 service on both the physical and logical hosts, and then execute the ps command to check whether JP1/AJS3 processes remain. If JP1/AJS3 processes remain, use the kill command to terminate them forcibly.
If the JP1/AJS3 service on the logical host cannot be started, execute the jajs_killall.cluster command on that logical host to terminate the remaining processes forcibly.
-
-
If the KAVS8033-E message (An error occurred during the processing of the connection source restriction function. (code: cause-code, host: host-name) maintenance-information) is output to the integrated trace log:
Reading of the connection permission configuration file might have failed. Confirm the following:
-
The connection permission configuration file is in the environment settings storage folder.
-
You have access permission for the connection permission configuration file.
-
-
If you try to start the JP1/AJS3 service without entering the IP address of the local host in the manager connection permission configuration file, the KAVU4335-E message (The request sent from the host (connection-source-IP-address) was ignored. (reason, host-name)) is output to the integrated trace log, and the JP1/AJS3 service stops. If this message is output, enter all IP addresses, including the loopback and logical host IP addresses, that might be used as the connection source IP address in the manager connection permission configuration file, and then try to start the JP1/AJS3 service.
(2) A JP1/AJS3 service takes too much time to start
When JP1/AJS3 starts, it requests the authentication server to perform initialization. Even if the authentication server is not running, JP1/AJS3 can still start, but startup takes time.
To prevent a slow startup, before you start JP1/AJS3, start the authentication server.