To access the Cluster Monitor page, select ToolsSystem InformationCluster Monitor.
The Cluster Monitor page provides access to tools that help you focus on your computing environment from the perspective of clusters and their attributes. You can view the results in three frames within the workspace; the left frame displays the Cluster Tree and the Status Messages list. The right frame displays selected items from the Cluster Tree, called the Data Viewer.
Three main graphical user interface (GUI) regions are used with the Cluster Monitor:
Cluster Service Names and Status Icons
HP Systems Insight Manager is designed to show the service name in parenthesis alongside the cluster name. For example, a cluster might look like the following when in tree view:
If a cluster service type is not pre-defined in HP Systems Insight Manager, the service name is replaced by (enterprise number-cluster type number). The enterprise number identifies the company that is providing the cluster service. The cluster type number is the cluster type number unique to that company. For example, Novell has a cluster type called Novell MSCS, where the cluster type number is 2301. If Novell MSCS is not defined in HP Systems Insight Manager, and it might be shown as:
where 23 is the enterprise number and 1 represents the MSCS product.
To find an enterprise number assigned to a particular company or vendor, go to http://www.iana.org/assignments/enterprise-numbers
The status of an attribute is also reflected in the color of the parent node and cluster icons, so changes in node attributes (CPU, System, or Disk resource) or cluster attributes (MSCS resource) are visible even when the tree is collapsed. When more than one cluster or node attribute is in an abnormal state, the parent cluster or node icon reflects the most severe state.
The status of clusters, cluster members, and resources as reported in the Cluster Monitor is independent of cluster and cluster member status shown on the system list pages. Because the status measurements are independent, status in the Cluster Monitor and status in a system list page can differ. For example, the Cluster Monitor might report cluster status as critical while the system list page shows status to be normal.
Like all nodes listed under the
System Lists region, each cluster, node, and attribute in the list is marked with a status message status icon. The icons have the following meanings:
Normal. The node or attribute is operating normally
Minor. A minor problem exists with the node or attribute
Major. A major problem exists with the node or attribute that should be addressed immediately
Critical. HP Systems Insight Manager can no longer communicate with the node or attribute
Unknown. HP Systems Insight Manager is not able to obtain management information about the node or attribute
The Cluster Monitor relies on software components called Cluster Resources to monitor specific cluster attributes. For example, a Disk resource that monitors Disk space and a System resource link to the HP Insight Management Agents page on the node.
When a resource detects a state change, it creates a status message in the Cluster Monitor and an HP Systems Insight Manager event. When you click a status message in the Status Messages list, the corresponding resource displays the state information in the Data Viewer. When the resource goes back to a normal state, the status message is deleted from the Status Messages list. The HP Systems Insight Manager event remains.
Resources are represented in the Cluster Tree as an attribute. Like clicking a status message in the Status Messages list, clicking an attribute in the Cluster Tree displays details in the Data Viewer.
Two classes of resources are distinguished by their scope:
Cluster-level resources monitor attributes on a cluster-wide basis (MSCS cluster status, for example). Attributes monitored by cluster-level resources are listed with nodes at the second level of the Cluster Tree.
Node-level resources monitor attributes on individual nodes (Disk or CPU utilization, for example). Attributes monitored by node-level resources display at the lowest level of the hierarchy.
Vertically. Drag the border between the Status Messages list and Cluster Tree to make the list longer or shorter. Lengthening the Cluster Tree shortens the Status Messages list.
Horizontally. Drag the border between the Status Messages list and the Data Viewer to make the list wider or narrower. The Status Messages list and Cluster Tree widen together, and widening them narrows the Data Viewer.
Expanding or Collapsing an Entry
Double-click an entry, or click
to the left of an entry to expand the sub-tree below that entry.
to the left of an expanded entry to collapse the sub-tree.
The Cluster Tree is an expandable hierarchy of a cluster, its nodes, and attributes. It is available from the System Lists region by selecting a cluster list and then clicking a cluster name on the cluster list page or from the Tools menu. Each cluster expands to show a list of nodes in the cluster. Each node expands to show a list of attributes for that node.
Click any entry in the Cluster Tree to display information about that entry in the Data Viewer.
The Cluster Tree status is a propagated status, which is dependent on the propagation of the items in its hierarchy. The status of an attribute or resource is reflected in the color of the parent node and cluster icons. Changes in node or cluster resources are visible even when the Cluster Tree is not expanded. When more than one cluster or node resource is in an abnormal state, the parent cluster or node icon reflects the most severe state because the resource statuses are propagated upward. The node status in the Cluster Tree is not tied to any node status in tables provided by the MSCS resource, only to the resource status for that node.
Status Messages list
The Status Messages list is on the left side of the Cluster Monitor page. Resources add entries, called status messages, to this list to bring your attention to cluster attributes that are in an abnormal state.
The Status Messages list header summarizes the total number of status messages in the list and the number of new status messages that have not been examined.
When a cluster enters an abnormal state, a corresponding status message is added to the Status Messages list so that you can track the status of the problem. Following the summary header, status messages are listed in order of descending priority (Critical, Major, Minor, Normal, and Unknown) and decreasing age (oldest, longest outstanding status message first, most recent status message last). The oldest Critical status message displays first, followed by the most recent Critical status message, then the oldest warning status message displays, followed by the most recent warnings, then the oldest Unknown status message displays followed by the most recent Unknown status messages.
Each status message is a link to a display of relevant information. Click a status message to display this information in the Data Viewer.
When you place the mouse pointer over a status message, a summary of the most recent state change related to the status message pops up. The pop-up includes a summary of the total number of problems subsumed under the status message.
Status Messages and Underlying Object States
One status message can represent more than one abnormal state. A status message reports the state of a cluster attribute on a node- or cluster-wide basis. Disk space, for example, is reported as a node-level attribute. If several Disks on a particular node are short on space, a single status message reports Disk problems on that node. Details on the number and identity of the critical Disks display in the Data Viewer area when you follow the status message link. The underlying problems related to a status message are displayed in the pop-up summary shown when the mouse pointer is placed over the status message.
Status Messages as Links to Problem Details
Each status message in the list is a link to detailed information about the problems that are causing the abnormal condition. Like all browser links, the color of the status message text indicates whether the link has been followed. Unlike typical browser links however, a status message reverts to the color of an unfollowed link when new details are available or a change of status occurs. A status message can represent multiple underlying problems. So, when a new problem is added to a status message that has already been examined, the text of the status message returns to the color of an unfollowed link to indicate that there is new, unexamined information relevant to the status message.
The cluster information provided in a Data Viewer is set by an HP Systems Insight Manager user with full configuration rights on the OptionsCluster MonitorCluster Settings and OptionsCluster MonitorNode Settings pages.
There are several ways to view detailed information related to a status message, cluster, node, or attribute. When you select a status message in the Status Message List or select a cluster, node, or attribute in the Cluster Tree, details are displayed in the Data Viewer. The information displayed in the Data Viewer depends on the item selected:
When you select a status message in the Status Messages list, the Data Viewer displays details about the status message.
When you select a node in the Cluster Tree List, the Data Viewer displays administrative information for the node.
When you select an attribute in the Cluster Tree List (Disk or CPU, for example), the Data Viewer displays attribute-specific information for the particular node or cluster where the attribute icon is located.
When you select a cluster in the Cluster Tree, the Data Viewer displays problem reporting and administrative information for the cluster.
For the cluster selected in the Cluster Tree View, the Data Viewer section lists the cluster identity, location, and contact information, and links to aggregate data display.
Some node-level resources summarize information for all the objects they monitor in an entire cluster. Click a link in the Aggregate Data area to see data for all objects in the cluster on a single page. Refer to Cluster Monitor - Aggregate Data View for more information.
For the node selected in the Cluster Tree view, the following displays: node identification, location, and contact information, along with other administrative information.
This Node Information area lists identification, location, and contact information for the node. Unless otherwise indicated, all values are entered on the Cluster Node Settings page. Refer to Clusters - Specifying Node Settings and Cluster Monitor - Node Fields.