NCM long discovery times in DX NetOps

Products

• Network Observability

Issue/Introduction

You experience extraordinarily long discovery times in NCM, accompanied by UI errors such as "Resource Not Found" and "Failed to retrieve all Jobs" ․․․․

SYMPTOMS:

• Discovery jobs take an excessive amount of time (e.g., 7.5 hours for 3 IP addresses).

 

Environment

25.4.x

Cause

•  This occurs when running discovery profiles with excessive device classes or inefficient discovery methods (SNMP/TFTP vs. SSH).

   

Resolution

• Administrator permissions within the NCM UI required

   

STEPS:

1. OPTIMIZE DISCOVERY PROFILES: Reduce the scope of discovery to improve performance:

   • Navigate to Auto Discovery Manager.

   • Edit the discovery profile and remove unneeded Device Classes.

   • Selecting only relevant device classes reduces discovery time exponentially.

2. UPDATE DISCOVERY MECHANISM: Use more efficient communication protocols:

   • Change the discovery mechanism from SNMP Sweep to Ping Sweep or SSH where applicable.

   • Verify that devices are using SSH for in-band management instead of SNMP/TFTP.

   • SSH typically responds faster than SNMP during discovery.

Additional Information

PREVENTION: Create specific discovery profiles for different device types rather than using a single profile with all device classes enabled.