hum - device availability threshold

Unanswered Question
Sep 7th, 2010
User Badges:

Hello,


I have lms 3.2 and i am using hum to probe devices (device availability template).

i use this setup to be alerted by email when one branch office goes down. i monitor

the main switch on each of these branch office. All of these offices are connected via 10 meg mpls

(good networkperformance)


I do receive some false positive alerts. Seems to be too much sensitive at times....


i have increased # violation to 2 (threshold criteria) and i received false alarms again..


i need some hints to make my setup more robust. i just want to receive an e-mail

when it's really down..


is there a more robust template to check device availability ?


if you can help me on this...

  • 1
  • 2
  • 3
  • 4
  • 5
Overall Rating: 0 (0 ratings)
Loading.
Joe Clarke Sat, 09/11/2010 - 17:35
User Badges:
  • Cisco Employee,
  • Hall of Fame,

    Founding Member

There's not much else you can do here.  HUM is polling sysUpTime to determine device availability.  SNMP being UDP may be getting dropped due to congestion.  DFM is typically used for alerting on device availability, but it also uses SNMP and ICMP.


Something else you might consider is using IPM to setup an IP SLA collector on one of your local devices to test reachability to your remote office.  While IP SLA can use ping, your alert criteria can be X of Y or consecutive.  You could also use a TCP connection instead to a node in your remote office (e.g. telnet or SSH to the remote router).

g.palavecino Mon, 09/13/2010 - 06:36
User Badges:

I have the same problem with HUM / threshold. Notify via email OID false positives with CRC.

Actions

This Discussion