Uptime resets randomly in the Management > Cluster page due to Cassandra restarts

Products

• VCF Operations

Issue/Introduction

In the VCF Operations for Logs 9 cluster, frequent uptime resets are seen in the Management > Cluster page.

The cluster is appropriately sized for the number of Events per second that are received - VCF Operations for logs Sizing Considerations

Environment

VCF Operations for Logs 9.0.x.0

Cause

Expensive queries are causing the internal Cassandra database to grow its memory heap usage until it hits the max limit. Once this max limit is reached, the OS detects that the available memory is not available for the heap to grow, so it kills the loginsight process. A watchdog service then restarts the loginsight service.

This restart of the loginsight service resets the uptime value.

Resolution

1. Log in the VCF Operations for Logs UI and navigate to Management > System Monitor > Active Queries
2. Click the Cancel button (X icon) for any long running queries listed

When using grouping in queries, select groupings that have fewer categories to prevent the issue from occurring.

Additional Information

Troubleshooting Query Results

Best Practices for Creating Search Queries

Cluster nodes uptimes reset randomly.