5.4.5 Flow of Problem Investigation and Response Based on Dashboard
This section explains the operation workflow from monitoring to investigation when using dashboards.
- Organization of this subsection
(1) Sequence of actions from IM management node status monitoring
The [Node Status] pane can be monitored in units of IM management node.
For example, if the entire system consists of eight systems, you can define eight node state monitors per system in the dashboard to monitor the entire system.
The procedure for monitoring is as follows.
-
Start monitoring.
-
When an error occurs, the color of the corresponding system panel changes.
-
Clicking the corresponding colored panel changes IM management node in the tree to the selected state.
-
In the [Event List] window, only events under the corresponding IM management node are filtered.
-
In the event list, check the "Event Detail" message of the relevant event, identify the cause, and investigate.
-
If the cause is related to OS resources or process performance, use dashboard to trend parse OS resources and investigate the cause.
(2) Series of operations from alert information monitoring
Alert-information can be monitored on a system-by-system or host-by-host basis, for example, in IM management node.
For example, if a system consists of 16 hosts, you can monitor alert information for the system by defining alert information monitoring on a per-host basis in the dashboard.
The procedure for monitoring is as follows.
-
Start monitoring.
-
When an alert occurs, the color of the Alerts panel for that host changes and displays the number of alerts occurring.
-
CClicking on the corresponding coloured pane displays a list of IM management node and the alerts that have occurred, and you can also see the time of occurrence. Clicking the displayed IM management node changes to the tree selection.
-
In the [Event List] window, only events under the corresponding IM management node are filtered.
-
From the event list, check the "Event Detail" message of the relevant event judging from the alert occurrence time, identify the cause, and investigate.
-
If the cause is related to OS resources or process performance, use the dashboard to trend parse OS resources and investigate the cause.