Submit a ticketCall us

Solarwinds & Cisco Live! Barcelona
Join us from the 29th of January to the 2nd of February at Cisco Live 2018 in Barcelona, where we will continue to show how monitoring the network with SolarWinds will keep you ahead of the game. At our booth (WEP 1A), we will demonstrate how SolarWinds network solutions can help. As a bonus, we are also hosting a pre-event webinar - Blame the Network, Hybrid IT Edition with our SolarWinds Head Geek™, Patrick Hubbard on January 24th - GMT (UTC+0): 10:00 a.m. to 11:00 a.m. There's still time to RSVP.

Home > Success Center > Archive > 2018Jan04 - Deletes > NCM nodes failing Connections after upgrade to NCM 7.3 or later

NCM nodes failing Connections after upgrade to NCM 7.3 or later

Table of contents
Created by Bill Corgey, last modified by Kevin.Swinson on Jan 04, 2018

Views: 692 Votes: 1 Revisions: 5


This article provides information on NCM not being able to connect to nodes after upgrading to NCM 7.3 or later.


  • NCM 7.3 through 7.4.1


Prior to NCM 7.3, the use of Additional Polling Engine was either not supported or was optional and in most installations not being used. This would mean NCM would work from the Primary NCM server for the connection to devices (either Primary Polling Engine or designated NCM server). 

A lot of the time an ACL was in place to permit the vty, tftp, ssh, telnet and so from the NCM server specifically and this could be accomplished with a single ip address

After the introduction of NCM 7.3, not only were the databases merged but also:

  • NCM if needing to be integrated with NPM had to be installed on the NPM server
  • If there were Orion Additional Polling Engines involved, the NCM poller plug-in is required
    • This meant that all NCM actions were now spread out to all the Additional Polling Engines and the connections from NCM to the device is initiated from the Additonal Polling Engine the device is assigned to.

If there was an ACL or even a Wwindows Firewall that was set up to permit the connection from NCM, it is likely the ACL and/or firewalls would need to be modified to allow connections from all engines involved or if NCM was moved to a different server.


Last modified