1

I am running a Nagios server on vmware 4.0 and every now and again during the day it alerts that some of servers cannot be reached via ICMP, clearly staging that a certain percentage of packets send are lost. This does not happen to all servers.

I know the servers are actually up, because I have done a parallel test from another windows box, just using simple ping and no packets were lost. I also now from the way we monitor out switch that no packets were lost on those particular ports.

Could anyone suggest a way to troubelshoot this further ? AT present this points to the nagios server itself loosing packets.

thanks

Ankh2054
  • 1,414
  • 13
  • 23
  • what are the warning/critical thresholds set to? are you using check_ping or check_icmp? – Keith Jul 24 '12 at 17:53

1 Answers1

4

The biggest cause of packet loss on VMs is failure to install the VMware tools package. Make sure that is installed and running always for all VMs.

ThatGraemeGuy
  • 15,473
  • 12
  • 53
  • 79
  • I think this is the right track to go down - I doubt Nagios or the nagios VM are losing the packets, I think it's more likely a problem with the virtualised server interacting with the network through the physical machine. – Bart B Jul 03 '12 at 10:43
  • thanks I will give that a go. I actually realised that it did not have the right version of vmware tools installed, so i've removed this - re-installed with correct version and installed the vmxnet drivers. Will give it 24 hours. thanks for the help :) – Ankh2054 Jul 03 '12 at 12:52
  • just had another few ICMP packets drop. Any other ideas ? Could it the vmware installation ? – Ankh2054 Jul 03 '12 at 13:28
  • 1
    Usually if you've confirmed all is well with the VMware tools, then normal network troubleshooting applies, i.e. physical cable checks, try another NIC, etc. – ThatGraemeGuy Jul 05 '12 at 13:12