Set up and monitor Cisco Unified Computing Systems (UCS)
Enhanced support for monitoring Cisco UCS devices was introduced in Orion Platform 2018.4, for VMAN 8.4 and SAM 6.8. The enhanced support was also added to NPM in NPM 12.5.
If you have NPM 12.5 and later, VMAN 8.4 or later, or SAM 6.8 or later, see Monitor Cisco UCS devices in the Orion Platform Administrator Guide.
NPM 12.4 and earlier
To gain an overview of all information provided by UCS in NPM 12.4 and earlier, add the parent UCS controller to the SolarWinds Orion database. Afterward, you can add connected UCS devices, such as fiber interconnects, chassis, rack servers, and blade servers, to create a UCS cluster.
- Confirm that LDAP authentication is not enabled on your UCS device. See the device documentation for details.
- Verify in the UCS console that the fiber connects have external IP addresses.
If the external gateway, external IP address, or external mask are set to 0.0.0.0, edit them with values valid for external devices.
- Add the parent UCS controller node to the database.
If the node shows up in the All Nodes list in italics or with ‘n/a’ as the state, click on it, and click Yes to add the device.
- Log into the Orion Web Console as an administrator.
- Click Settings > Manage Nodes, and then click Add Node.
- On the Add Node page, provide the IP Address, or provide the host name and select Dynamic IP Address.
- Select an ICMP polling method and provide credentials.
To gather maximum UCS data, select Most Devices: SNMP and ICMP as the Polling Method.
- Select the UCS Manager Credentials check box.
- Provide UCS credentials information, including the port on which the UCS Manager listens:
- Use HTTPS
- User Name
- UCS Password
- Click Test to verify credentials, and then click Next.
- On the Choose Resources page, select the resources and statistics you want to monitor, and then click Next.
- Advance to the Add Pollers and Change Properties pages to add relevant pollers and adjust node properties.
- Review your information, and click OK, Add Node.
For Fabric Interconnects, the polling method must be SNMP. The SNMP version is the customer's preference.
- To select the proper view we use the existing View By Device Type feature.
- To ensure that Standard Poller does not overwrite MachineType and other fields, we use EntityType to identify UCS node in the Standard Poller (and so force Standard Poller not to overwrite our required fields). This same mechanism is also used for the ESX VMWare API.