Hi,
I still have 2 requests concerning lockd. Could you please tell me if these
features will be accepted, because if not I won't go on to code it, as I
don't want to do the patches if they will never be accepted.
1) client side:
have a nlm_prog or lockprog as option for mount, which says which rpc
program number to use with lockd.
2) server side:
Monitor the caller_name instead of the rpc peer.
These 2 thinks are needed by sec_rpc (at least I think so) to have lockd
tunneling with more than one server, and working statd.
Pat
_______________________________________________________________
Sponsored by:
ThinkGeek at http://www.ThinkGeek.com/
_______________________________________________
NFS maillist - [email protected]
https://lists.sourceforge.net/lists/listinfo/nfs
>>>>> " " == Patrice DUMAS <[email protected]> writes:
> 1) client side:
> have a nlm_prog or lockprog as option for mount, which says
> which rpc program number to use with lockd.
> 2) server side:
> Monitor the caller_name instead of the rpc peer.
> These 2 thinks are needed by sec_rpc (at least I think so) to
> have lockd tunneling with more than one server, and working
> statd.
I disagree totally. They are merely workarounds in order to try to
shoehorn NFS into a model for tunnelling that doesn't fit the bill...
Tunnelling should be done by setting up a proper virtual
point-to-point interface. That would do the job without needing any of
these hacks.
Cheers,
Trond
_______________________________________________________________
Don't miss the 2002 Sprint PCS Application Developer's Conference
August 25-28 in Las Vegas - http://devcon.sprintpcs.com/adp/index.cfm?source=osdntextlink
_______________________________________________
NFS maillist - [email protected]
https://lists.sourceforge.net/lists/listinfo/nfs
Hi,
> I disagree totally. They are merely workarounds in order to try to
> shoehorn NFS into a model for tunnelling that doesn't fit the bill...
Indeed.
> Tunnelling should be done by setting up a proper virtual
> point-to-point interface. That would do the job without needing any of
> these hacks.
I agree again, but the point in tunneling rpc based services through
ssh like with sec_rpc, is that it is quite easy to set up, almost off the box.
Maybe there exists point-to-point solutions as easy to set up but I don't
know them.
Pat
_______________________________________________________________
Don't miss the 2002 Sprint PCS Application Developer's Conference
August 25-28 in Las Vegas - http://devcon.sprintpcs.com/adp/index.cfm?source=osdntextlink
_______________________________________________
NFS maillist - [email protected]
https://lists.sourceforge.net/lists/listinfo/nfs
>>>>> " " == Patrice DUMAS <[email protected]> writes:
>> Tunnelling should be done by setting up a proper virtual
>> point-to-point interface. That would do the job without needing
>> any of these hacks.
> I agree again, but the point in tunneling rpc based services
> through ssh like with sec_rpc, is that it is quite easy to set
> up, almost off the box. Maybe there exists point-to-point
> solutions as easy to set up but I don't know them.
In that case, if there is sufficient demand for it, perhaps this would
be a project worth investigating in its own right? I must admit that I
for one am still not entirely sure which tool is best for the job: I
once suggested the TUN/TAP interface, but perhaps there are better
candidates?
The point is that we don't want to have to reinvent the wheel for NFS,
then SMBFS, CIFS, <insert flavour-of-the-month-networking-filesystem
here>. In particular if this will involve unnecessary messing about
with new incompatible APIs for the NFS utilities and mount programs,
then I for one don't want to know. At least not until somebody has
seriously looked into how much work the business of "doing it right"
would involve.
Cheers,
Trond
_______________________________________________________________
Don't miss the 2002 Sprint PCS Application Developer's Conference
August 25-28 in Las Vegas - http://devcon.sprintpcs.com/adp/index.cfm?source=osdntextlink
_______________________________________________
NFS maillist - [email protected]
https://lists.sourceforge.net/lists/listinfo/nfs
On Fri, Jun 14, 2002 at 02:05:22PM +0200, Patrice DUMAS - DOCT wrote:
>
> I agree again, but the point in tunneling rpc based services through
> ssh like with sec_rpc, is that it is quite easy to set up, almost off the box.
> Maybe there exists point-to-point solutions as easy to set up but I don't
> know them.
>
It is fairly easy to set up a PPP link over ssh using the pppd "pty"
command, provided you have root access at both ends for doing the
initial configuration. Then you have full IP connectivity. In
addition, it is possible to do routing for connecting all the networks
at both sides (I'm using something similar for years now).
--
Frank
_______________________________________________________________
Don't miss the 2002 Sprint PCS Application Developer's Conference
August 25-28 in Las Vegas - http://devcon.sprintpcs.com/adp/index.cfm?source=osdntextlink
_______________________________________________
NFS maillist - [email protected]
https://lists.sourceforge.net/lists/listinfo/nfs