Determine Monitoring Requirements

Before you configure monitoring in a new environment, use the following guidelines to determine the monitoring requirements of the environment.
Before you configure monitoring in a new environment, use the following guidelines to determine the monitoring requirements of the environment.
To determine the monitoring requirements of the environment, complete the following steps:
After you determine the monitoring requirements, configure monitoring.
Identify Data Sources
During the installation, you register the data aggregator as a data source. Many implementations of
DX NetOps Performance Management
include other data sources, such as
DX NetOps Spectrum
CA Application Delivery Analysis
. To import inventory from other data sources, register those data sources.
Plan Discovery
Discovery is the process that
DX NetOps Performance Management
uses to build an inventory of devices in the network. You can define SNMP profiles with the authentication credentials to communicate with devices in the network and use discovery profiles to limit discovery.
Plan Collections and Monitoring Profiles
Collections are system groups that group devices for monitoring. Monitoring profiles control how often to poll devices and which metrics to collect. Associating a collection with a monitoring profile causes
DX NetOps Performance Management
to monitor the devices according to the parameters in that profile. You can identify a basic set of monitoring profiles and collections up front to reduce maintenance overhead.
Plan Metrics
You can assign metric families to each monitoring profile to specify the metrics that the system should collect. Consider the metrics that are most useful for monitoring the environment.
You might plan to collect the following metric families:
  • Interface
  • CPU
  • Memory
  • Disk
  • QoS
Plan Poll Rates
In each monitoring profile, you can specify how frequently to poll for specific metrics.
You might plan to collect metrics at the following poll rates:
  • Interface (1 minute)
  • Interface (5 minutes)
  • CPU (5 minutes)
  • Memory (5 minutes)
  • Disk (5 minutes)
  • IPSLA (5 minutes)
  • QoS (5 minutes)
Plan Monitoring Profiles
You can create a matrix like the following example to determine the necessary monitoring profiles:
Polling Rate
1 minute
5 minutes
15 minutes
You can then use the planned metrics and poll rates to plan the monitoring profiles.
In this example, the administrator plans the following monitoring profiles:
  • Fast Interfaces (includes Interface metrics that are polled at 1 minute)
  • Standard (includes Interface, CPU, Memory, and Disk metrics that are polled at 5 minutes)
  • IPSLA (includes IPSLA metrics that are polled at 5 minutes)
  • QoS (includes QoS metrics that are polled at 5 minutes)
You can use monitoring profile filters to refine which managed items the monitoring profile applies.
Plan Collections
Consider each of the managed items as you plan the collections.
In this example, the administrator plans the following collections:
  • Fast Interfaces (including WAN links) for association with the Fast Interfaces monitoring profile
  • Active Devices (including all active devices with some exceptions) for association with the Standard monitoring profile
  • Core Routers (including routers in core) for association with the IPSLA monitoring profile
  • Distributed Routers (including routers in the distributed network) for association with the QoS monitoring profile
You can use group rules to refine which managed items the collection applies.
Plan Monitoring Profile Filters
Monitoring profile filters specify criteria that governs which components are monitored. Only the component items that match the filter criteria are polled for the associated metric family. Filtering limits SNMP traffic and ensures that the system monitors only relevant components. The filters of each monitoring profile are assessed independently.
Plan Group Rules for Collections
Use rules to keep the collections up-to-date when systems and networks change. Newly discovered items that meet rule specifications are added to collections. If existing items no longer meet rule requirements, they are removed from collections. After you create a rule, you can modify it by deleting filters or adding subrules.
Plan Events
An event is a message that provides information about what is happening in
DX NetOps Performance Management
. Events provide information for monitoring the health and status of your system and your environment. All events include basic information, such as related devices and the time of the occurrence that triggered the event.
For more information, see Events.
Plan Threshold Profiles
Threshold profiles trigger events when specified conditions occur in associated groups. Event rules define the conditions that trigger events. Each event rule is set to a single metric family, and determines the conditions that cause or clear a violation. Each threshold profile requires at least one event rule.
Plan Monitoring Profile Event Rules
For thresholds that apply broadly to devices in the network, you can add event rules at the monitoring profile level. For example, a rule that creates an event whenever CPU utilization exceeds 95 percent could apply to any device.
Plan Dashboards
Dashboards contain sets of views that show you polled data as meaningful information.
DX NetOps Performance Management
includes several out-of-the-box dashboards that provide basic information about your infrastructure. To set up dashboards that match your specific monitoring requirements, create a custom dashboard or edit an out-of-the-box dashboard.
Plan Reports
You can access and share On-Demand reports, which dynamically retrieve the most recent data sets from specific sets of items or groups. You can also access and share dashboards and views.