Submit a ticketCall us

ebook60.pngHow to be a Cisco® ASA ace

Our eBook, Thou Shalt Not Pass…I Think?! can help you overcome the challenges of monitoring and managing Cisco ASA firewalls. This eBook is a great read if you’ve been frustrated with monitoring firewalls, managing ACL configs, and troubleshooting VPN connections.

Get your free eBook.

Home > Success Center > User Device Tracker (UDT) > UDT Business Layer Process has high CPU usage on an additional polling engine

UDT Business Layer Process has high CPU usage on an additional polling engine

Updated 21st Sept 2017

Overview

An additional polling engine is at 95% CPU usage. The Business Layer was split to run separate processes, the the UDT business layer is at approximately 50-70% CPU usage.

Environment

  • UDT 3.3

Cause 

This seems to have been caused by the load on that server -- a combination of the number of jobs, ports, and large devices, and possibly SNMP delays.

 

  • UDT Scalability
    • UDT now supports:
      • 90K ports per polling engine 
      • 300K ports or more, maximum per instance (multiple polling engines) 
  • UDT_Job:
    • 300 Jobs L2 and L3 Jobs over last 3 days
    • Bringing back 70,000 ports

Resolution

  1. Edit UDT Settings -> Layer 2 Polling and change the value to 1200 minutes.
  2. Edit UDT Settings -> Layer 3 Polling and change the value to 1200 minutes.
  3. EditUDT Settings -> SNMP pacing delay time and change the value to 1 (currently it is set 0 by default) on Device Tracker Advanced Settings page.
    If there are undesired side effects you may set this value back to zero.
 

 

Last modified

Tags

Classifications

Public