Operating System - OpenVMS
cancel
Showing results for 
Search instead for 
Did you mean: 

Numerous calls to icmp_error

 
SOLVED
Go to solution
roose
Regular Advisor

Numerous calls to icmp_error

Hi Folks,

I'm seeing a lot of this errors on our nodes and would just like to get some points on how to troubleshoot here:

$ tcpip show prot icmp
icmp:
28116 calls to icmp_error
0 errors not generated because old ip message was too short
0 errors not generated because old message was icmp
Output histogram:
echo reply: 362556
destination unreachable: 28116
0 messages with bad code fields
0 messages < minimum length
0 bad checksums
0 messages with bad length
Input histogram:
echo reply: 887448
destination unreachable: 224
echo: 362556
router solicitation: 3
address mask request: 3
362556 message responses generated
$

1. Is this kind of error specific to a single network card? Because on our nodes, we have 2 network cards and 1 is not used but is still active.
2. I have read somewhere that this might be due to the fact that the router where our node is connected to might be experiencing high utilization. What should I ask our network admin to check first in this case?
3. The item "destination unreachable" seems to equate to the number of calls to icmp_error. How can I see which destination it is trying to reach? Will this be visible on any of our log files? I can't seem to see it on the operator.log yet.

This node is an ES80 on OpenVMS 7.3-2, TCPIP v5.4 ECO7.

Thanks.
1 REPLY 1
Jim_McKinney
Honored Contributor
Solution

Re: Numerous calls to icmp_error

The "destination unreachable" typically indicates that your adjacent router is having a problem forwarding packets from your server onward to their destination. Routers send these messages back to the source when it can not successfully forward a packet. Could be that the destination host is down, a network link on the other side of the router is down, the host that is supposed to recieve the packets is not listening on the target port, or a myriad of other things. You _probably_ need not be concerned with your host. You'll _likely_ want to be looking on the other side of your local routers for the issue (if there still is an issue).