Problem with rpc_ping times, permission denied mounts, etc.

Problem with rpc_ping times, permission denied mounts, etc.

[Chris Feist]

Ian,

Looking at Sun's documentation it doesn't appear that they take into 
account ping times when using automounter maps.  It looks like they just 
sort by network proximity and then weight. Would it make sense to remove 
the rpc_ping requests and only use network proximity and weight.

Also, it seems like it would be best to return all of mounts from the 
get_best_mounts function in order from best to worst.  Then, when it's 
time to mount the directories we can try the first and if we get 
permission denied or it isn't there we can fail over to the second, etc. 
  As opposed to what happens now if the best host has nfs temporarily 
disabled (or the wrong permissions set) the whole directory fails to mount.

Thanks,
Chris

Re: Problem with rpc_ping times, permission denied mounts, etc.

[Mike Waychison]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Chris Feist wrote:
> Ian,
> 
> Looking at Sun's documentation it doesn't appear that they take into
> account ping times when using automounter maps.  It looks like they just
> sort by network proximity and then weight. Would it make sense to remove
> the rpc_ping requests and only use network proximity and weight.
> 

IIRC, Solaris 'network proximity' is defined from closest to farthest:

1) Localhost
2) On the same subnet
3) rpc_ping times

> Also, it seems like it would be best to return all of mounts from the
> get_best_mounts function in order from best to worst.  Then, when it's
> time to mount the directories we can try the first and if we get
> permission denied or it isn't there we can fail over to the second, etc.
>  As opposed to what happens now if the best host has nfs temporarily
> disabled (or the wrong permissions set) the whole directory fails to mount.
> 
> Thanks,
> Chris
> 
> _______________________________________________
> autofs mailing list
> autofs@linux.kernel.org
> http://linux.kernel.org/mailman/listinfo/autofs


- --
Mike Waychison
Sun Microsystems, Inc.
1 (650) 352-5299 voice
1 (416) 202-8336 voice

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
NOTICE:  The opinions expressed in this email are held by me,
and may not represent the views of Sun Microsystems, Inc.
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.2.5 (GNU/Linux)
Comment: Using GnuPG with Thunderbird - http://enigmail.mozdev.org

iD8DBQFCAUINdQs4kOxk3/MRAkpBAJ0cEsM3IEkmgT5EhmNfyoNZ8TsY2QCgjCxe
+gJM4Zbhkv1FZapcpRplvyM=
=DKAX
-----END PGP SIGNATURE-----

Re: Problem with rpc_ping times, permission denied mounts, etc.

[Ian Kent]

On Wed, 2 Feb 2005, Chris Feist wrote:

> Ian,
> 
> Looking at Sun's documentation it doesn't appear that they take into 
> account ping times when using automounter maps.  It looks like they just 
> sort by network proximity and then weight. Would it make sense to remove 
> the rpc_ping requests and only use network proximity and weight.
> 
> Also, it seems like it would be best to return all of mounts from the 
> get_best_mounts function in order from best to worst.  Then, when it's 
> time to mount the directories we can try the first and if we get 
> permission denied or it isn't there we can fail over to the second, etc. 
>   As opposed to what happens now if the best host has nfs temporarily 
> disabled (or the wrong permissions set) the whole directory fails to mount.
>

Sounds good.

We probably need to assign priorities in some sort of semi-democratic 
way, to these "larger than bug fix" tasks.

Ian