2M ping packets

[John Owens_2]

Experts,
Does anyone know of any known issues with HP servers in dealing with large ping packets?

Regards,
John

[Ron Kinner · Accepted solution]

Do you mean "Ping of Death" size packets? Not since 10.20.

http://216.239.53.100/search?q=cache:hnASmJm784oC:www.pp.asu.edu/support/ping-o-death.html+%22ping+of+death%22&hl=en&lr=lang_en|lang_de&ie=UTF-8

However, 11.0 does have a problem with pings of all sizes. It often responds with a source quench instead of a reply. There is a patch for it but you can just turn it off in ndd. (don't forget to edit /etc/rc.config.d/nddconf or it will go away at the first reboot).

Ron

[John Owens_2]

Thanks Rob,
I don???t believe that the problem encountered is the Ping O??? Death, as the problem arises with packets that are much smaller than 65510. The reason this is a problem for us is cause we have an application that relies on mounted file system for communications. The application is not working correctly. When we tested out the connectivity the problem with a large ping became apparent. At first we were blaming the network itself but we found that other servers hanging off the same router are accepting large pings without any problems. I will try editing /etc/rc.config.d/nddconf. What exactly does this file do?

Regards,
John

[melvyn burnard]

nddconf sets configuration parameters for the ndd command, used to "tune" the network.
It replaces the nettune command used in 10.X
What systems/applications are you seeing these issues on?
I recall long ago that with NFS we had issues with Novell servers not liking packet sizes (or something like that) and we had to mount them with reduced sizes.
Just a thought.

[John Owens_2]

Melvin,
One of my Co-workers has just informed me that last night they twigged nddconf and it seems to have resolved out issue.

As for your question the application that I am supporting is just a network monitoring system that relies on a mounted file system for communication between redundant servers.


Many Thanks for your reply,
John