18.5.1 Problems in Windows or UNIX
- Organization of this subsection
(1) A large number of JP1 events occur within a short period of time, causing a delay in registration and transfer.
When an error generates a large number of JP1 events in quick succession, the JP1 event reporting the error might take a long time to appear in JP1/IM - View. This might delay the execution of any JP1/AJS jobs triggered by these JP1 events. Forwarding of the JP1 events being processed when the error occurs resumes when the failed host is restored and the event service restarts.
If the occurrence of large numbers of events causes too much load on the system or affects job processing, it might be absolutely necessary to stop the JP1 events from being forwarded. In such cases, use the event-forwarding suppression command (jevagtfw) to suppress forwarding of the JP1 events. For details on event forwarding suppression using the jevagtfw command, see 2.5.3 Using a manager to suppress event forwarding from an agent with large numbers of JP1 events. For details on the jevagtfw command, see jevagtfwin 15. Commands.
If the event-forwarding suppression command (jevagtfw) cannot be used in your environment, initialize the event database on the event server from which the JP1 events were forwarded. If you want to save the JP1 events already registered in the event database, use the jevexport command to output the event database in CSV format before you initialize the database.
For details on the procedure to initialize the event database, see 10.2 Initializing the event database.
To prevent the generation of large numbers of JP1 events, adjust the JP1 event forwarding conditions in the forwarding settings file (forward). You can also set a threshold to detect large numbers of events, and automatically suppress event forwarding. For details on the threshold-based suppression of event-forwarding, see 2.5.4 Setting a threshold to automatically suppress forwarding of large numbers of events.
(2) The event database is corrupt.
The event database might be corrupt due to the following reasons:
-
Sudden loss of power because of a power outage or some other reason
-
Backing up or restoring the event database by an OS command or backup software while the event service is active
-
Editing the event database in a text editor
-
Redirecting command output results or other information to the event database
-
A hard disk error
Even if the event database is damaged, the event service will still start or continue running, and new JP1 events can still be registered or acquired as normal. However, damaged records will not be retrieved or acquired. Damaged records in the event database might also affect the performance of event searches from JP1/IM - View.
The KAJP1057-W, KAJP1058-W, or KAJP1059-E message is output to the event log, syslog, and integrated trace log when the event database is damaged. Initialize the event database if any of these messages appears.
For details on the procedure to initialize the event database, see 10.2 Initializing the event database.
(3) The event database is deleted from the source event server, resulting in delays in JP1 event registration on the target event server.
If the event database is lost from the source event server due to, for example, server replacement or disk failure, and the event database serial number is reset to 0 on the source event server, the source event database serial number of a forwarded JP1 event becomes smaller than the maximum value for the source event database serial number of the source event server that is recorded in the duplication prevention table on the target event server. Therefore, when a JP1 event is forwarded, an event database search is performed on the source event server to check for duplications. This might result in delaying JP1 event addition on the target event server. If the event database is lost from the source event server for a reason such as server replacement or disk failure, recover the event database by using the procedure described in the following subsections:
If the source event server is not available for operation, but you have added a JP1 event for clearing the duplication prevention table (user event) on the target event server, you can clear the maximum value for the source event database serial number of the source event server in the duplication prevention table.
- Procedure:
jevsend -i 7FFF8001 -m "DPT CLEAR EVENT" -e OBJECT_NAME=source-event-server-name
- Notes
-
-
This procedure is applicable only to file-format duplication prevention tables.
-
If the duplication prevention table is re-created by the jevdbmkrep command when no JP1 events have been received from the source event server since a JP1 event for clearing the duplication prevention table was issued, the maximum value for the source event database serial number of the source event server is not cleared. After you have executed the jevdbmkrep command, issue a JP1 event for clearing the duplication prevention table again.
-
(4) JP1/Base does not function as defined in a definition file.
Because some of the settings for a service running on a host do not reflect the settings in the service's definition file, the service might not function as expected. In order to determine the source of the problem, you must compare the operating information of the service that is currently running and the contents of the service's definition file.
You can use the jbsgetopinfo command to view the operating information of a service that is currently running. This command provides operating information defined in the forwarding settings file for the event service, the action definition file for the log file trap, and the action definition file for the event log trap (Windows only) for the service. Compare the operating information with the contents of each respective definition file. If there is a difference, take the necessary action to correctly incorporate the contents of the definition files. For details on the jbsgetopinfo command, see jbsgetopinfo in 15. Commands.