Chuck Lever wrote:
Why not clientaddr? That is already used for NFS version 4, and it works
just fine for version 3 as well. I can't think of any reason to have
a new variable.
That may make sense for NFSv2/v3, but nfs(5) documents the clientaddr
specifically as the client's NFSv4 callback server address. Is there
ever a case where we want to specify a NFSv4 callback address that's
different from the source address of your NFSv4 forward channel? Are we
sure these will always be the same?
Conversely, do we always want to limit our NFSv4 traffic to a single
bind address when clientaddr= is specified?
IMO the coding and documentation overhead for reusing clientaddr would
be the same or greater than creating a new mount option, and the result
would be more confusing. These two options do different things. I'd be
interested to hear opinions from others on the list.
"bindaddr=" might be more clear than "localaddr=". I checked Solaris
10, but it doesn't seem to have a similar capability. So, no guidance
there...
Either way is fine with me. I'll change it to be bindaddr unless
I hear another opinion.
Thanks,
Ben
--
Ben Greear <greearb@xxxxxxxxxxxxxxx>
Candela Technologies Inc http://www.candelatech.com
--
To unsubscribe from this list: send the line "unsubscribe linux-nfs" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
