The only time I've received this error is when I've killed the socket
using NETSTAT *CNN
I would imagine that ENDTCP would also trigger this error, but I don't
have the luxury of experimenting with the ENDTCP command :)
On Thu, 27 Jun 2002, David Gibbs wrote:
> Folks:
>
> Anyone know what might be causing a tcp/ip read() function to set an errno
> of EUNATCH (3448)?
>
> This is happening sporadically, with no discernable pattern.
>
> The message that we're retrieving from the strerror function is "the
> protocol required to support the specified address family is not available",
> but all the TCPIP docs I can find reference EUNATCH as "Protocol driver not
> attached" (might mean the same thing, but it's inconsistent).
>
> Thanks!
>
> david
>
> p.s. This post may duplicate, the first one I sent didn't go through.
>
