14.9 Considerations for JP1/IM system-wide maintenance
Among the hosts monitored by JP1/IM, some might be running processing that cannot be halted even when the server requires maintenance. For example, if a host runs 24 hours a day, 365 days a year, you will need to schedule maintenance work so as to minimize its impact, by migrating job processing to another server or by suspending some processing for a short time, for example.
During maintenance work, you might need to stop and start the server several times, or halt job processing temporarily, potentially issuing a large volume of events that detected these start/stop operations as errors.
To enable continuous operation monitoring by JP1/IM, you must filter out redundant events and keep only those that are required. You should therefore consider the order in which maintenance tasks will be carried out, and perform settings in advance to filter out unwanted JP1 events.
This section describes some points to consider in maintaining the entire system including JP1/IM, and discusses how to implement a maintenance schedule that allows for uninterrupted system monitoring. It also provides an example of carrying out urgent maintenance when a problem occurs on an agent.
- Organization of this section