5.7.1
Table Of Contents
- VMware vCenter Operations Manager Getting Started Guide
- Contents
- VMware vCenter Operations Manager Getting Started Guide
- Introducing Custom User Interface Features and Concepts
- Introducing Common Tasks
- Logging In and Using vCenter Operations Manager
- Monitoring Day-to-Day Operations
- Handling Alerts
- Optimizing Your Resources
- Designing Your Workspace
- Working with Dashboards
- Create a Dashboard
- Clone a Dashboard
- Edit a Dashboard
- Delete a Dashboard
- Create a Dashboard Template
- Hide a Dashboard Tab
- Change the Order of Dashboard Tabs
- Delete a Dashboard Template
- Configure Dashboard Switching
- Share a Dashboard
- Stop Sharing a Dashboard
- Export a Dashboard
- Import a Dashboard
- vSphere Dashboards
- Working with Widgets
- Working with Dashboards
- Using and Configuring Widgets
- Edit a Widget Configuration
- Supported Widget Interactions
- Configure Widget Interactions
- Advanced Health Tree Widget
- Alerts Widget
- Application Detail Widget
- Application Overview Widget
- Configuration Overview Widget
- Custom Relationship Widget
- Data Distribution Analysis Widget
- Generic Scoreboard Widget
- GEO Widget
- Health Status Widget
- Health Tree Widget
- Health-Workload Scoreboard Widget
- Heat Map Widget
- Mashup Charts Widget
- Metric Graph Widget
- Metric Graph (Rolling View) Widget
- Metric Selector Widget
- Metric Sparklines Widget
- Metric Weather Map Widget
- Resources Widget
- Root Cause Ranking Widget
- Tag Selector Widget
- Text Widget
- Top-N Analysis Widget
- VC Relationship Widget
- VC Relationship (Planning) Widget
- Define Metric Sets for a Widget
- Index
Alert Types and Subtypes
vCenter Operations Manager generates several types of alerts, and each alert type has its own triggers and
contents. An alert might be caused by abnormal behavior of one or more metrics for a resource, including a
tier or application, or when a fingerprint predicts an upcoming problem.
By default, vCenter Operations Manager generates all types of alerts except for classic abnormality alerts.
Because they do not involve KPIs, classic abnormality alerts are unlikely to require any action.
Smart Alerts
A Smart Alert is triggered when an internally calculated value indicates that a problem occurs, or soon will
occur. The internally calculated value can be a correlated prediction of future behavior or a dynamic
threshold breach on a KPI. Except for metrics that are designated as KPIs, Smart Alerts do not depend on
user-defined values. Smart Alerts have certain subtypes.
Table 2‑5. Smart Alert Subtypes
Subtype Description Determined By
KPI DT Breach A KPI breached one of its internally
calculated dynamic thresholds.
Analytics based on a user-defined
KPI.
KPI Prediction vCenter Operations Manager predicts that a
KPI will soon breach a threshold. The current
combination of metrics might match a stored
metric fingerprint that predicts the breach, or
a metric that is correlated with the KPI has
breached its threshold.
KPI Prediction alerts are always critical.
Analytics based on a user-defined
KPI.
Early Warning vCenter Operations Manager correlation
algorithms calculate a greater than 90 percent
chance above the noise threshold that there is
a problem with an application.
Early Warning alerts are always critical.
Analytics based on application
topology.
Classic Alerts
A classic alert is an alert that is generated by classic monitoring software. It relies on user-defined hard
thresholds, not vCenter Operations Manager dynamic thresholds or other advanced calculations. Classic
alerts have certain subtypes.
Table 2‑6. Classic Alert Subtypes
Subtype Description Reason
KPI HT Breach A user-defined KPI breached a user-
defined hard threshold.
A user-defined KPI breached one of its
user-defined hard thresholds.
Abnormality A non-KPI metric breached one of its
internally calculated dynamic
thresholds.
A single non-KPI metric breached one of its
dynamic thresholds.
Notification A non-KPI metric breached a user-
defined hard threshold. Typical system
and network management applications
send this type of alert.
A single non-KPI metric breached one of its
internally calculated dynamic thresholds.
Administrative Alerts
An administrative alert indicates a problem with vCenter Operations Manager, the monitoring software
from which it collects data, or the enterprise network. Administrative alerts have certain subtypes.
VMware vCenter Operations Manager Getting Started Guide
26 VMware, Inc.