Multiple Adapter are Automatically Changing to Warning State and Aria Operations is Slow/Unresponsive

Products

• VCF Operations/Automation (formerly VMware Aria Suite)

Issue/Introduction

Per the VMware Aria Operations Sizing Guidelines, Aria Operations nodes and cloud proxies must have sufficient CPU and memory allocated based on the number of objects and metrics collected in the environment. As the environment grows, the amount of vCPUs and memory may need to be increased on the nodes and cloud proxies.

Determine the number of objects and metrics collected by reviewing the System Audit:

1. Log into Aria Operations and go to Administration > Control Panel > Audit.

2. Click on the System Audit tab.

If your cluster is undersized, you may experience the following symptoms:

• Errors messages appear on the console of the nodes similar to: 

  NMI watchdog: BUG: soft lockup

• The status for some nodes may show "Waiting for analytics"

• Aria Operations UI is slow or unresponsive

• Data collection will fail intermittently.

• Multiple adapters may go into Warning state intermittently.

Environment

• Aria Operations 8.18.x

Cause

• An undersized Aria Operations cluster leads to performance degradation and intermittent data collection failures.

Resolution

To address data collection and performance issues stemming from inadequate sizing, increase the CPU and/or memory allocation.

When increasing CPU or memory for Aria Operations nodes, only perform the following steps on Analytic nodes (Primary, Primary Replica, Data).

Note: All Aria Operations nodes in a cluster must have the same CPU and Memory configuration.

To add additional CPUs and/or Memory:

1. Log into the Admin UI as admin and click the Take Cluster Offline button to take the cluster offline. Do not proceed unless the cluster is offline. See: VMware Aria Operations Cluster Management

2. Log into the VMware vSphere client as an administrator.

3. Right-click the desired Aria Operations node or cloud proxy and click Power > Shut Down Guest OS.

4. Right-click the powered off node or cloud proxy and click Edit Settings.

5. Set the CPU and Memory (GB) values to the appropriate values based on the sizing guidelines.

6. Click OK.

7. Right-click the node and click Power On.

8. Repeat steps 3-7 on all other analytic nodes in the cluster.

9. Log into the Admin UI as admin and click the Bring Cluster Online button to bring the cluster online.