The monitoring profile determines which metric families are collected for all devices in associated device collections and how often polling occurs. The metric families in the monitoring profiles determine which components are added to the system for devices in assigned device collections and which metrics are collected for those components.
Monitoring profiles are global across all tenants.
The relationship of monitoring profiles to device collections governs monitoring. Monitoring can be triggered in the following ways:
Consider the following best practices for creating monitoring profiles:
Before you configure monitoring profiles, see Determine Monitoring Requirements.
Copy default monitoring profiles to create customer-specific monitoring profiles. Remove the collections that are applied to the default monitoring profiles. This model makes it easier to customize and manage device polling.
Note: Do not modify the DA Health Fast, Normal, and Slow monitoring profiles. These profiles are used for self-monitoring of the CA Performance Management components.
To specify which metrics to gather from which devices and components, create monitoring profiles. For example, create a custom Systems monitoring profile and configure the profile to poll resource utilization counters only from SNMP.
Monitoring profiles are available in all tenant work spaces. For example, create a "Service Router Monitoring" monitoring profile and use the profile for all tenants. You do not have to create a separate "Service Router Monitoring" monitoring profile for each tenant. Device collections that are associated with the monitoring profiles are tenant-scoped.
Tip: To create basic monitoring profiles, copy the out of the box monitoring profiles. Custom profiles provide flexibility.
Only Administrator users in the default tenant can modify monitoring profiles. In an MSP environment, all changes to monitoring profiles apply to all tenants.
Follow these steps:
When you use the 60-minute rate for an existing device, a No Data To Display message appears in views with a time range of Last Hour. If you change the dashboard setting to a prior hour, you can see earlier data. However, the view does not display the latest data until the new poll cycle completes.
Important! One-minute polling on large device collections or devices with many components can affect the performance of the system.
Note: To poll critical interfaces at a faster rate, use a separate monitoring profile with a filter. For more information, see Poll Critical Interfaces Faster than Non-critical Interfaces.
Specify the change detection rate. The change detection rate determines how often the system identifies changes to component items on monitored devices. If a device is associated with multiple monitoring profiles, change detection occurs as the fastest specified rate. To determine the configuration changes, CA Performance Management uses reconciliation algorithm that is defined in the metric family.
(Optional) To automatically start monitoring newly detected components, select Automatically Update Metric Families. To manually control when changes to monitoring occur, clear this option.
Note: If you clear this option, the system does not automatically apply changes to monitoring. The Events Display dashboard displays configuration events for the detected changes. To apply the changes to the devices, go to the Data Aggregator administration menu, Monitored Devices, and open the Polled Metric Families tab. Select the appropriate metric family, and click Update Metric Family.
When monitoring changes are applied to devices, components that are no longer detected on the devices are marked "Not Present".
Assign metric families to the Selected Metric Families list.
Tip: By default, all metrics in each selected metric family are collected. To select a limited set of metrics, see Configure Metric Filtering.
Event rules set conditions to trigger Threshold Violation events. Each rule is based on a single metric family and contains criteria to raise or clear the violation. The event rules apply to each component on all devices in collections that are assigned to the monitoring profile.
You cannot add event rules to factory monitoring profiles. Only Administrator users in the default tenant can modify event rules on monitoring profiles.
Tip: For best performance and granular control, configure event rules on threshold profiles instead of monitoring profiles. For more information, see Configure Threshold Profiles.
Follow these steps:
Select a custom monitoring profile.
Click the Event Rules tab.
Click New to create a rule, or click Edit to modify an existing rule.
For detailed information about creating event rules, see Configure Threshold Profiles.
Events are raised and cleared when the devices in associated collections satisfy the event rule conditions.
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.
Important! CA Performance Management monitors device components if any of the monitored components pass the filter criteria of any associated monitoring profile.
Note: Monitoring profile filters apply after discovery. The system creates the component items but does not send SNMP requests to monitor the items. Monitored components are reassessed according to the change detection interval. Vendor certification filters prevent the creation of component items that do not match the filter criteria. Vendor certification filters might not recognize changes at the component level. Both filter types reduce SNMP traffic to the same level. Use vendor certification filters only when the monitoring profile filter cannot filter component item.
You want to monitor all interfaces with 5-minute polling, and one type of interface with 1-minute polling. One monitoring profile has a poll rate of 5 minutes. This monitoring profile is associated with a device collection. A second monitoring profile has a poll rate of 1 minute. The interface metric family on that profile has a filter that specifies the type of interface to poll at a 1-minute interval. This monitoring profile is also associated with the same device collection. CA Performance Management polls the specified interface type at a 1-minute rate and all other interfaces at a 5-minute rate.
When you have the same devices that are associated to two monitoring profiles, the two profiles must have filter criteria. In this scenario, effectively, you combine the two filters by stating exclusion criteria in both. Otherwise, the monitoring profile without the filter causes all components on the device to be polled. Therefore, Filter 2 should appear as follows:
Only Administrator users in the default tenant can add or edit monitoring profile filters.
Follow these steps:
To apply the polling behavior of a monitoring profile to devices, associate the profile with a device collection. When multiple monitoring profiles are associated with a device collection, Data Aggregator uses the fastest specified poll rate.
Important! Do not associate monitoring profiles with the All Devices device collection. Doing so can cause extra SNMP requests being made to pingable-only devices, and can result in sporadic metric family support.
Metric families such as QoS, MPLS, and various Response Path Test metric families can contribute to significant SNMP requests. For more information about the implications and restrictions of SNMP requests to your network devices, see the vendor documentation or contact the vendor.
Administrators and Tenant Administrators can modify associations between monitoring profiles and device collections. Administrators see only the associated device collections for the current tenant.
Follow these steps:
Follow these steps:
Factory monitoring profiles let you start monitoring devices in the environment quickly. Factory monitoring profiles are automatically associated with factory device collections. For example, the Router monitoring profile is associated with the All Routers device collection.
The following factory monitoring profiles can have a significant performance impact: