Job Management Partner 1/Automatic Job Management System 3 Command Reference 1
Format
jpomanevreset
[-h logical-host-name]
[-F service-name]
[-a execution-host-name|-all]
[-s]
[-e]
Description
Deletes the information that is to be held by the event action control manager. The command also deletes information on the execution host after communicating with the execution host specified in the -a option.
When a large number of events related to start conditions occur, event action control is in a state of high load, with the result that a slowdown or similar problem may occur in the entire system. However, you can execute this command to return the system to normal operation from the next JP1/AJS3 startup. If the system has several agents, you should use the jpomanevshow command in advance to identify the agent that has caused a large number of events.
You can execute this command only when the JP1/AJS3 host service is running and the target JP1/AJS3 scheduler service has stopped. If you execute the command when the target JP1/AJS3 scheduler service is running or the overall JP1/AJS3 service has stopped, the command results in an error.
For details about how to stop the scheduler service, see jajs_spmd_stop in 2. Commands.
Execution privileges
In Windows: Administrators privileges
In UNIX: Superuser privileges
Arguments
-h logical-host-name
Specifies the name of the logical host responsible for event reset when you are working with a cluster system.
You can specify a string of 1 to 255 bytes.
If you omit this option, the logical host name specified in environment variable JP1_HOSTNAME is assumed. If the environment variable JP1-HOSTNAME is not specified, the local host name is assumed.
-F service-name
Specify the name of the scheduler service for which you want to delete information.
You can specify a character string of 1 to 30 bytes.
If you specify the -a option together with this option, you can limit the jobs for which you want to delete information to the event jobs and jobnets with start conditions executed on the host specified in the -a option. If you do not specify the -a option, the command deletes the information about all jobs executed from the scheduler service specified in the -F option.
If you omit this option, the system assumes the default scheduler service name.
-a execution-host-name|-all
When you perform an event reset, you can delete only information about a specific target host by specifying the host name in the -a option. The execution host name you specify in this option is the execution host name specified in the -s option of the ajsagtadd command. For example, if an execution host name in FQDN format is specified in the -s option of the ajsagtadd command, you need to specify in the -a option the execution host name in FQDN format.
You can use the ajsagtshow command to check the execution host name for the execution agent name.
Example: ajsagtshow -l
For details about the ajsagtadd command, see ajsagtadd in 2. Commands. For details about the ajsagtshow command, see ajsagtshow in 2. Commands.
You can specify a character string of 1 to 255 bytes.
If you do not want to limit the target hosts on which you want to delete information about jobs executed from the scheduler service specified in the -F option, specify -all.
If you omit the -a option or if you specify the -a option without specifying the -F option, the target host for the default execution agent is assumed.
-s
Deletes only the information on the manager without communicating with the execution host specified in the -a option.
Specify this option if any of the following conditions is met. If you omit this option under the following conditions, it may take time before the command is terminated.
- When the JP1/AJS3 service on the execution host is cold-started
- When communication between the manager and the execution host cannot be performed properly because of communication failure
If you omit this option, the information on the execution host is also deleted after communicating with the execution host specified in the -a option.
-e
Deletes only unprocessed data while continuing to execute the event job and start conditions. Because the monitoring state on the execution host is terminated once before re-execution, there is a time difference from when the JP1/AJS3 service is started until actual monitoring starts on the execution host.
If you omit this option, the event job and start conditions being executed are terminated.
When using this option to execute the command, you need to perform the steps shown in the table below:
Table 2-21 What you need to do before and after starting JP1/AJS3
Timing Action Before starting JP1/AJS3 On the execution host, remove the cause of the problem. After starting JP1/AJS3 On JP1/AJS3 - View, check the state of the job and re-execute the command, as necessary.
Notes
- Always separate an option and its value by at least one space character.
- You cannot execute several instances of this command at the same time. Wait for one command to be completed before you move on to another command.
- If the target host specified in the -a option has been defined as the target host for multiple execution agents, information about these execution agents is also deleted.
- The command does not check whether the target host specified in the -a option has been registered as the target host for the execution agent in agent management information. Therefore, if JP1/AJS3 has not been installed on the specified target host, it could take time before the command terminates.
- If you execute the command when the folder to be used for storing information about the event action control manager, which is used by the event action control manager, does not exist, the system operates as if there were no data. The command is terminated normally. For details about the folder for storing information about the event control action manager, see A.1 JP1/AJS3 - Manager files and directories in the manual Job Management Partner 1/Automatic Job Management System 3 Troubleshooting.
- Note that the command will terminate normally even if you fail to delete the file or send data to the execution host. If the KAVT8342-E A file access error occurred. (file-name) (maintenance-information) or KAVT8343-W An attempt to communicate with the agent has failed. (agent-host-name) (maintenance-information) message is output in the standard error output file, you must manually delete the file according to the specified procedure.
- If you fail to delete files, the KAVT8342-E A file access error occurred. (file-name) (maintenance-information) message for each of the unsuccessful files is output to the standard error output file. If you fail to delete a large number of files, a large number of error messages are output accordingly. This implies that you cannot view all the messages. For this reason, you should redirect the standard error output file when executing this command.
- Example
- jpomanevreset 2> file-name
- If the KAVT8343-W An attempt to communicate with the agent has failed. (agent-host-name) (maintenance-information) message is output, the information about the agent involving the error is not deleted. As necessary, therefore, you should cold-start the JP1/AJS3 service on the execution host.
- If the -all option is specified, the command reports that the specified scheduler service has been cold-started to the target hosts on which the status of event jobs and jobnets with start conditions is Now running. In addition, the command also performs the same report to the target hosts for which unreported information exists. If the -a option is specified, the command reports that the specified scheduler service has been cold-started to the specified target host even if the status of event jobs and jobnets with start conditions is not Nor running.
- If an error such as a network error occurs while data is being sent to the target host, the behavior of the command is as follows:
For this reason, if the target host or the JP1/AJS3 service on the target host has stopped, it could take time before the command terminates. If the target host name cannot be resolved into the IP address, the command also retries acquisition of the IP address three times at one-second intervals.
- When the agent host cannot be connected (a connection attempt times out)
The command does not retry sending of data.
- When another network error occurs
The command retries to connect (up to two attempts) at 10-second intervals.
- You execute this command in the same environment as that of JP1/AJS3 operation. For example, suppose that you changed DNSEstablish after a JP1/AJS3 stop. Executing this command changes the host name used when communicating with the execution host, so that you cannot identify which manager information needs to be deleted on the execution host. This may cause a problem; for example, the information on the execution host will not be deleted, as expected.
- If there are many files under the folder used for storing information about the event control action manager, it will take time before the command is completed because of the presence of many files that are to be searched for and deleted.
- If you specify the local host name in the -h option, the system will operate by assuming that a physical host is specified.
- If you execute the command after specifying the -e option, the system continues to monitor event jobs and start conditions but deletes the event information and event start information generated before the event reset. For this reason, you may see Now running for the status of a job in which an event occurred before an event reset or Now queuing for the state of a job in which monitoring has already started. If this occurs, for Now running jobs, you must re-generate events meeting the conditions. For Now queuing jobs, meanwhile, you should either use them as they are or terminate them and then re-register them.
- On the target host side, only the information managed by the target host is deleted. Files remaining as trash are not deleted. If you want to delete files left on the target host as trash, cold-start the JP1/AJS3 service on the target host side.
- If you execute the command without specifying the -e option, monitoring of event jobs and start conditions ends. If monitoring of these is necessary, perform registration again. You can ignore the KAVS1400-E and KAVS0265-E messages output to the integrated trace log because these messages pertain to termination of monitoring of event jobs and start conditions.
Return values
0 Normal end. 1 There is an argument improperly specified. 2 Initialization failed.
There is an error in the definitions or environment settings on the logical host.3 Privileges not sufficient for task requested. 4 The scheduler service specified in the -F option is running. 5 The same command is being used simultaneously. 100 A file access error occurred. 101 Memory shortage occurred. 255 A system error occurred.
Additional information
The status of JP1/AJS3 - View displayed when JP1/AJS3 is normally started is different from that displayed when JP1/AJS3 is started after you use the jpomanevreset command to recover the status of the event action control manager. The state of JP1/AJS3 - View is also different depending on whether you specify the -e option in the jpomanevreset command: that is, whether the event jobs and start conditions are to be continued. These different statuses are compared and summarized below:
- When you hot-start JP1/AJS3 (-hot)
Target Status Status before JP1/AJS3 is stopped Normal state After executing the jpomanevreset command Being executed on the local host Being executed on the remote host Specifying -e Not specifying -e Being executed on the local host Being executed on the remote host Being executed on the local host Being executed on the remote host Event job (in jobnet) Now running Ended abnormally #1, #2 Now running #2 Ended abnormally/Unknown end status Now running/Unknown end status Unknown end status Wait for prev. to end Wait for prev. to end #3 Wait for prev. to end #3 Being held Being held #3 Being held #3 Now queuing Now running/Ended abnormally Now running Ended abnormally/Unknown end status Now running/Unknown end status Unknown end status Other statuses No change No change Start condition Now monitoring Now monitoring Now monitoring Monitor terminated Other statuses No change No change Event job in start condition Now running Now running Now running/Unknown end status Unknown end status Now queuing Now queuing Now queuing/Unknown end status Other statuses No change No change
- #1
- The Now running status occurs when the system is restarted after a system down.
- #2
- The Unknown end status may occur depending on the timing of the termination processing.
- #3
- If the Not executed + Ended status occurs when the preceding unit is handled as an abnormal end.
- When you warm-start JP1/AJS3 (-warm)
Target Status Status before JP1/AJS3 is stopped Normal status After executing the jpomanevreset command Being executed on the local host Being executed on the remote host Specifying -e Not specifying -e Being executed on the local host Being executed on the remote host Being executed on the local host Being executed on the remote host Event job (in jobnet) Now running Unknown end status Unknown end status Wait for prev. to end Not executed + Ended Not executed + Ended Being held Not executed + Ended Not executed + Ended Now queuing Unknown end status Unknown end status Other statuses No change No change Start condition Now monitoring Now monitoring Now monitoring/Monitor terminated Monitor terminated Other statuses No change No change Event job in start condition Now running Now running Now running/Unknown end status Unknown end status Now queuing Now queuing Now queuing/Unknown end status Other statuses No change No change - When you cold-start JP1/AJS3 (-cold)
Target Status Status before JP1/AJS3 is stopped Normal state After executing the jpomanevreset command Being executed on the local host Being executed on the remote host Specifying -e Not specifying -e Being executed on the local host Being executed on the remote host Being executed on the local host Being executed on the remote host Event job (in jobnet) Now running Not registered Not registered Wait for prev. to end Being held Now queuing Other statuses Start condition Now monitoring Other statuses Event job in start condition Now running Now queuing Other statuses - If the startup mode is not specified when JP1/AJS3 is started
The status of each item when the startup mode of the scheduler service is used is the same as that of the corresponding item shown above.
Example
See the Example in jpomanevshow in 2. Commands.
Copyright (C) 2009, 2010, Hitachi, Ltd.
Copyright (C) 2009, 2010, Hitachi Solutions, Ltd.