How to add many Pingable items at one time in Performance Management

Products

• CA Infrastructure Management
• CA Performance Management - Usage and Administration
• DX NetOps

Issue/Introduction

We need to bulk add around many devices to DX NetOps Performance Management.

We only want to monitor the devices for ICMP Ping, not for any SNMP even though the devices are SNMP capable.

Environment

All supported Performance Management releases

Resolution

Options to accomplish this goal?

1. The correct way to address this scenario is the use of ACLs or Firewall rules in the network.
   • Create Rules that prevent successful SNMP communications between Data Collector(s) (DC) and target devices.
   • Ensure the rules only allow Ping via ICMP between the DC(s) and target devices.
2. The way to address this via the Performance Management tool is via normal Discovery with a few special steps.
   • NOTE: This relies on never running a Discovery Profile against these Pingable items devices using a Discovery Profile tied to an SNMP Profile.
   • If that SNMP Profile has credentials that would succeed in communicating via SNMP with the target devices it would be updated to a ManageableDevice.
   • Once this is done only deleting and rediscovering as Pingable again would resolve that.
   • There is no demoting a device from ManageableDevice to Pingable only. An ACL or firewall rule per option 1 would prevent this possibility.

To execute the product method via Performance Management complete the following steps.

1. Select an SNMP Profile for use in this task.
   1. Ensure the SNMP Profile specified is configured with purposefully incorrect credentials. Credentials that would result in SNMP communication failure from the target devices.
   2. If necessary create a new one with the purposefully incorrect credentials for this use.
   3. In the SNMP Profile used engineering recommends changing the SNMP profile variables to 0 retries with a 100 ms timeout.
      1. It'll speed along the SNMP attempts and failures making the discovery cycle for this effort faster.
      2. Would need to use REST services on the Data Aggregator to make that change.
      3. The change can be made only after the SNMP Profile is created and synced over to the DA.
      4. Documentation with instructions for making the changes are found here in the Modify the Timeout and Retries Parameters section.
2. Create a new Discovery Profile. Configure it as follows:
   1. IPs/Hosts Tab
      1. Using a list if individual targeted IP addresses? Create a text file listing the IP addresses.
      2. Import the IP Address list file.
      3. If using a set ot IP Ranges, enter them in the UI.
   2. Schedule Tab
      1. Set schedule as needed, if needed.
   3. SNMP Tab
      1. Select the "Use specific list of assigned SNMP profiles".
      2. Specify the specific SNMP Profile selected or newly created from step 1 above.
   4. Advanced Tab
      1. Set Naming Order as needed.
      2. Ensure both "Use ICMP" and "Create Pingables" are enabled by checking their boxes.
3. Run the Discovery Profile.

When the Discovery Profile completes there is no requirement to create a new or unique Collection for these new Pingable device items. They will be placed by default in the out of the box default 'All Devices' Collection. That Collection is by default tied to the Reachability Metric Family.

Additional Information

Are the target devices going to be assigned an SNMP profile? Will Performance Management, on it's own, continue trying to contact the devices via SNMP?

Once discovered as a Pingable it will be known as an AccessibleDevices device type. These can only be tied to the Reachability Metric Family. These will never have SNMP communication attempts retried unless someone runs their IP addresses through a new Discovery Profile run.

Once that happens, if it is then promoted to an SNMP manageable device, it would need to be marked as a ManageableDevice device type. This is what would trigger the assignment to an SNMP profile. From there it gets added to Collections and tied to their associated Monitoring Profiles that discover SNMP based Certifications. That's only done when we're able to successfully communicate with a device via SNMP.