6.2.8 Selection of agent configuration or remote monitoring configuration
Using IM Configuration Management, you can use JP1/Base to monitor hosts in agent configuration, or remotely monitor hosts in a remote monitoring configuration.
In remote monitoring, a communication failure related to functional errors might cause log monitoring to stop or events to be lost. If the system cannot tolerate temporary stoppage of log monitoring, install JP1/Base and use it for log monitoring.
Use the following table to help decide whether to use monitoring using JP1/Base or remote monitoring.
Item |
Monitoring using JP1/Base |
Remote monitoring |
---|---|---|
You perform monitoring that exceeds the restrictions for remote monitoring. |
Y |
N |
The stoppage of log monitoring and loss of events is undesirable in the system.#1 |
Y |
N |
Log monitoring is required even while JP1/IM - Manager is stopped.#2 |
Y |
N |
Monitored hosts are frequently stopped. |
Y |
N |
WMI/NetBIOS (NetBIOS over TCP/IP) or SSH cannot be used on the system.#3 |
Y |
N |
Other than above. |
Y |
Y |
- Legend:
-
Y: Monitoring is performed.
N: Monitoring is not performed.
#1: When monitoring using JP1/Base, if a network failure occurs, JP1/Base acquires logs during the network failure. With remote monitoring, the monitoring stops if the failure cannot be recovered by retrying the operation.
#2: With remote monitoring, monitoring of log files and event logs that are output while JP1/IM - Manager is stopped is not supported. Therefore, for example, if remote monitoring is executed on JP1/IM - Manager in a cluster environment, messages issued during failover cannot be monitored.
#3: With remote monitoring, settings for remote communication must be configured on the host on which JP1/IM - Manager has been installed and on the monitored hosts. The communication method depends on the OS of the monitored host. WMI/NetBIOS (NetBIOS over TCP/IP) must be configured if the monitored host is a Windows host, or SSH must be configured if the monitored host is a UNIX host.
For details about remote monitoring, see 6.6 Managing remotely monitored hosts.
Some limitations apply to remote monitoring. For the limit values, see the following table.
Item |
Limit |
---|---|
Maximum number of remote monitoring units#1 |
1,024 |
Maximum size of the log files that can be remotely monitored |
64 MB |
Maximum size of log data that can be collected in one monitoring interval#2 |
10 KB |
#1: Total value of the number of log files monitored by remote log file trapping and the number of servers monitored by remote event-log trapping
#2: The total size of the differences between the collected log files and the previous log files.
For the functional differences between the agent configuration and the remote monitoring configuration, see the following table.
Item |
Agent configuration |
Remote monitoring configuration |
---|---|---|
Use |
|
When reducing load in the construction of an environment and for other operations has priority over monitoring accuracy#1 |
Monitored logs and file formats |
|
|
Scale |
|
|
Monitoring interval |
|
|
Event issued |
|
|
Number of log files that can be monitored in each monitoring operation |
|
32 |
Size of log files that can be monitored |
2 gigabytes or less |
64 megabytes or less |
Size of log data that can be collected during a monitoring interval |
Unlimited |
10 kilobytes or less#3 |
Operation when manager stops |
If the manager host stops or a network failure occurs between the manager host and a monitored host, an agent host can convert the event into a JP1 event, and retry transfer of the generated JP1 event to the manager host. (By default, transfer is retried at intervals of 60 seconds for a period of 3,600 seconds.) |
The log files and Windows event logs generated while JP1/IM - Manager is stopped cannot be monitored. |
Operation when network fails |
If network operation is restored after a failure before the maximum number of retries is reached and monitoring stops, the manager host will be notified of the log data collected during the failure period. Notification will be done only when none of the following restrictive conditions exist:
|
#1: If the size of the log data generated during the monitoring interval exceeds 10 kilobytes for a monitored log, JP1 events cannot be displayed in JP1/IM - View. If this is a possibility, consider using log monitoring in an agent configuration.
#2: When IM Configuration Management is used, the maximum number of log file traps that can be managed is 100 for each agent host.
#3: If the log data size exceeds 10 kilobytes, log information is not displayed in JP1/IM - View. Instead, a JP1 event (event ID: 00003FC5 or 00003FD5) is issued to notify the user. If a JP1 event is issued, directly reference the log files and event logs on the remotely monitored host to check the meaning of the error.
For details about other differences, see 6.6.7 Difference between agent configuration and remote monitoring configuration.