N.3 Adding hosts to the monitoring target during scale out
If an agent host is added during scale out, a JP1 event is sent from the agent host to the manager host. The manager host then performs the automated action corresponding to this JP1 event and adds the agent host to the system's monitoring targets.
|
|
![[Figure]](GRAPHICS/ZU991300.GIF)
- Organization of this subsection
(1) Reporting a scale out by using JP1 events
The first time the OS starts up after a scale out occurs, a scale-out notification event (event ID: 1011) is issued as a JP1 event to notify the manager host of the scale out. For details about this event, see 17.3.1(1) Details about event ID 00001011.
If an attempt to send the event fails, the OS will continue to attempt to send the event at intervals of 600 seconds for a maximum of six times.
The manager host to which scale-out notification events are to be sent must be specified beforehand in the script definition file. You can change the notification event's attributes (event ID and message).
For details about the script definition file, see N.8 Auto-scaling linkage script.
(2) Network settings and monitoring settings
The automated action for the scale-out notification event involves the configuration of network settings and monitoring settings as follows:
-
The host name and IP address of the agent host are added to the manager host's hosts file.
-
The agent host is added to the health check definition file (jbshc.conf), and then JP1/Base is reloaded to apply the definition.
-
The agent host is added to the configuration definition file (jbs_route.conf), and then the configuration information is distributed.
If JP1/PFM is used for monitoring, an alarm table is bound to the agent based on the alarm table name specified in the script definition file. You can specify the name of the alarm table to be bound to the agent in the script definition file. For details about the script definition file, see N.8 Auto-scaling linkage script.
(3) Checking whether monitoring was performed successfully
You can determine whether monitoring was performed successfully by checking the termination code of the automated action for the scale-out notification event. If the automated action terminated normally, the termination code will be 0. If the action terminated abnormally, the termination code will be a nonzero value.
If the automated action terminated abnormally, remove the cause of the error, and then re-execute the automated action.