0

I am using the check_host_alive command to send 5 packets every minute to all my routers at remote locations. I noticed today I received a notification from The AT&T Global Client Support Center that a router was down (which can take 5-30 minutes to send these notices out) and never received a notice from Nagios. I went onto Nagios and it is was showing the host as alive with a latency of 0ms. This tells me it is seeing the automated response from my router in the data center that, "TTL expired in transit" as a reply from the remote router. Is there anyway for me to tell nagios to check where the reply is comming from? I feel like other people have to of had this issue... I tested it with the check_ping command and it produced the same results. I have the command defined has %hostname% and the proper IP in the host definition, and it works fine for telling me the latency is high. Any ideas are welcome, I have already exercised my Google skills with no results.

EDIT:

root@IM-UBTU:/# /usr/local/nagios/libexec/check_ping -H 192.168.250.1 -w 100.0,10% -c 200.0,20% -vvv
CMD: /bin/ping -n -U -w 10 -c 5 192.168.250.1
Output: PING 192.168.250.1 (192.168.250.1) 56(84) bytes of data.
Output: From 10.69.10.2 icmp_seq=1 Time to live exceeded

It knows something is wrong why doesn't it give me a warning?

1
  • Edit with output of your check_ping command along with -vvv for verbosity
    – brent
    Feb 23, 2011 at 17:09

2 Answers 2

1

I wanted to try and provide a solution here as there weren't any answers.

Your best bet is to use a different service check for what you want, unfortunately I don't know enough programming to provide a modified check_ping plugin.

An alternative would be to change the check to something like telnet or ssh, not ideal but would provide you with a better response than ping.

For how to change this see my post here: Change how Nagios determines if server is offline?

1
  • Switched over to snmp for this. But I am accepting this since a modified check_ping plugin probably would have worked if I told responses from 10.69.10.1 are consider down... Wish I the time to do it though.. Jul 15, 2011 at 16:13
1

You can specify hops in your ping syntax after all of the switches, right before you specify the destination host. I did a little test on my network and it seems as if you will get no reply if the hop is unreachable.

This does seem like a pretty major flaw in check_ping, you should report it as a bug.

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .