14.9.2 Example of JP1/IM system-wide maintenance
The following describes an example of implementing maintenance throughout the JP1/IM system.
In this example, maintenance is implemented for the system hierarchy (IM configuration) managed by JP1/IM as shown in the following figure (one integrated manager, two base managers, and four agents).
|
This system monitored by JP1/IM is configured as defined by the JP1/Base configuration management. For details about defining a system hierarchy, see 9.4.3 Managing the system hierarchy.
- Organization of this subsection
(1) Order of maintenance
Maintenance is carried out in order, from the higher-level hosts to the lower-level hosts, following the definition of the system hierarchy. In this example, the order of maintenance is as follows:
-
Integrated manager (mgr1)
-
Base manager (m1)
-
Base manager (m2)
-
Agents (a1 and a2)
-
Agents (a3 and a4)
Figure 14‒21: Order of maintenance
Maintenance is carried out on the integrated manager (mgr1), base manager (m1 or m2), and agent (a1, a2, a3, or a4), in that order. This prevents redundant events from being acquired and required events from being overlooked.
- Important
-
Always schedule maintenance work on the manager (in this case, mgr1) first of all. Determine the order for the other hosts according to the system hierarchy.
(2) Settings on the integrated manager (JP1/IM - Manager) side
Using the JP1/IM - Manager functionality, set up filtering of redundant events and specify JP1 event acquisition while JP1/IM - Manager is in stopped state.
(a) Setting the event acquisition start location
Use the jcoimdef command to set the event acquisition start location. There are two settings for acquiring events from before JP1/IM - Manager restarts:
-
jcoimdef -b -1
Acquires JP1 events going back to the status when JP1/IM - Manager was last stopped.
-
jcoimdef -b xxx (xxx: length of time; from 1 to 144 (hours))
Acquires JP1 events going back the specified length of time before JP1/IM - Manager restarts.
There is also a setting for starting event acquisition from the JP1 events issued after JP1/IM - Manager starts, but the setting does not apply in this example because the starting point needs to be specified explicitly.
For details about setting the event acquisition start location, see 14.7.10 Setting the event acquisition start location. For details about the jcoimdef command, see jcoimdef in Chapter 1. Commands in the JP1/Integrated Management 3 - Manager Command, Definition File and API Reference.
(b) Setting common exclusion-conditions in an event acquisition filter
You can define common exclusion-conditions in an event acquisition filter to temporarily prevent some JP1 events from being monitored or exclude them from automated-action execution You can add and edit common exclusion-conditions from the Event Acquisition Conditions List window, leaving the standard filter conditions unchanged.
Maintenance is to be carried out in the order shown in Figure 14-21 Hierarchical structure of a system monitored by JP1/IM, starting with the integrated manager (mgr1), then the base managers (m1 -> m2), and lastly the agents (a1, a2 -> a3, a4). Four common exclusion-conditions separate from the event acquisition filter normally used, that is one for each setting, will be required for maintenance purposes. These common exclusion-conditions are described in the table below. Here it is assumed that the settings to prevent redundant events generated by a host undergoing maintenance from being collected or exclude such events from automated-action execution will be performed on the integrated manager (mgr1) only.
Common exclusion-condition ID |
Host undergoing maintenance |
Common exclusion-condition |
---|---|---|
1 |
Base manager (m1) |
Include source host m1 |
2 |
Base manager (m2) |
Include source host m2 |
3 |
Agents (a1 and a2) |
Include source hosts a1 and a2 |
4 |
Agents (a3 and a4) |
Include source hosts a3 and a4 |
Source host in the event acquisition conditions means a host from which events are issued.
If you want to set up the same type of filtering for the base managers, set common exclusion-conditions in event acquisition filters, referring to the settings given above. Note that common exclusion-conditions must be changed to extended mode if you want the conditions to exclude JP1 events from action execution. And if you use common exclusion-conditions in extended mode, you will be able to specify the application period of a condition group. For details about common exclusion-conditions for the event acquisition filter, see 4.2.7 Common exclusion-conditions.
For detailed settings for common exclusion-conditions, the setting windows differ depending on whether extended mode is used. See the following descriptions in the JP1/Integrated Management 3 - Manager GUI Reference:
-
Common exclusion-conditions: 3.15 Common Exclusion-Conditions Settings window
-
Common exclusion-conditions (when extended mode is used): 3.16 Common Exclusion-Condition Settings (Extended) window
(3) Settings in JP1/Base on the base managers and agents
Set the event forwarding retry parameters, bearing in mind how long the destination host will be in stopped state. Set the parameters on the base managers (m1 and m2) and agents (a1, a2, a3, and a4). By default, the retry time limit is 3,600 seconds (that is, JP1/Base retries at set intervals for 60 minutes). Adjust this value if JP1/Base on the destination host will be stopped for longer than 60 minutes.
For details, see the JP1/Base User's Guide.
(4) Maintenance procedures
The following describes the maintenance procedure on each host.
(a) Maintenance procedure on the integrated manager
Follow these steps to carry out maintenance on the integrated manager (mgr1):
-
Stop JP1/IM - Manager.
-
Stop JP1/Base.
-
Perform maintenance of JP1/IM - Manager and JP1/Base.
For details about backup and recovery of JP1/IM - Manager settings information and disk space management, see Chapter 1. JP1/IM System Maintenance in the JP1/Integrated Management 3 - Manager Administration Guide. For details about JP1/Base maintenance, see the description of maintenance in the JP1/Base User's Guide.
-
Start JP1/Base.
-
Start JP1/IM - Manager.
The same procedure applies for maintenance of JP1/IM - Manager on the base managers.
(b) Maintenance procedure on the base managers and agents (JP1/Base)
To carry out maintenance on the base managers (m1 and m2) and agents# (a1, a2, a3, and a4), switch the common exclusion-conditions associated with the event acquisition filter on the integrated manager (mgr1) to eliminate redundant events.
#: This applies when JP1/Base is used as agent.
The figure below shows the order of the maintenance tasks and when to switch the common exclusion-conditions defined for the event acquisition filter.
|
You can switch the common exclusion-conditions of the event acquisition filter using the jcochfilter command, the System Environment Settings window, or the Event Acquisition Conditions List window.
- Procedure for switching the common exclusion-conditions of the event acquisition filter
-
See 6.5.3 Switching the event acquisition filter to be applied in the JP1/Integrated Management 3 - Manager Administration Guide.
- Details on the jcochfilter command
-
See jcochfilter in Chapter 1. Commands in the JP1/Integrated Management 3 - Manager Command, Definition File and API Reference.
- Details on the System Environment Settings window
-
See 3.11 System Environment Settings window in the JP1/Integrated Management 3 - Manager GUI Reference.
- Details on the Event Acquisition Conditions List window
-
See 3.14 Event Acquisition Conditions List window in the JP1/Integrated Management 3 - Manager GUI Reference.
The maintenance procedure and files needing to be backed up will depend on the products installed on the base managers and agents. When carrying out maintenance, see the JP1/Base User's Guide or the documentation for the particular products that issue JP1 events.
- Note
-
When you use the jcochfilter command, you can switch the event acquisition filter automatically at a specified time, based on the JP1/AJS scheduling and calendar functionality. This also lets you change the monitoring status of the host automatically.