excluding some pattern to prevent useless mount request

excluding some pattern to prevent useless mount request

[Guillaume Rousse]

While looking in server's log, I found many such mount requests:
Nov 26 04:23:19 yquem rpc.mountd: refused mount request from
chouchen.inria.fr for /home/yquem/.hidden (/): no export entry
...
Nov 26 04:22:33 yquem rpc.mountd: refused mount request from
medoc.inria.fr for /home/yquem/.directory (/): no export entry

I guess they are caused by graphical file manager trying to find
metadata directories when displaying mount points. Is there a way to
prevent them ?

My current map content is:
# /, auto.home.yquem, autofs, village.inria.fr
dn: cn=/,ou=auto.home.yquem,ou=autofs,dc=village,dc=inria,dc=fr
objectClass: top
objectClass: automount
cn: /
automountInformation: -fstype=nfs,hard,intr,nosuid,nfsvers=3
graves:/home/yquem/&

Re: excluding some pattern to prevent useless mount request

[Ian Kent]

On Mon, 2006-11-27 at 17:40 +0100, Guillaume Rousse wrote:
> While looking in server's log, I found many such mount requests:
> Nov 26 04:23:19 yquem rpc.mountd: refused mount request from
> chouchen.inria.fr for /home/yquem/.hidden (/): no export entry
> ...
> Nov 26 04:22:33 yquem rpc.mountd: refused mount request from
> medoc.inria.fr for /home/yquem/.directory (/): no export entry
> 
> I guess they are caused by graphical file manager trying to find
> metadata directories when displaying mount points. Is there a way to
> prevent them ?

This looks like a problem with hald.
Try updating it.

Ian

Re: excluding some pattern to prevent useless mount request

[Jeff Moyer]

==> On Mon, 27 Nov 2006 17:40:54 +0100, Guillaume Rousse <Guillaume.Rousse@inria.fr> said:

Guillaume> While looking in server's log, I found many such mount requests:
Guillaume> Nov 26 04:23:19 yquem rpc.mountd: refused mount request from
Guillaume> chouchen.inria.fr for /home/yquem/.hidden (/): no export entry
Guillaume> ...
Guillaume> Nov 26 04:22:33 yquem rpc.mountd: refused mount request from
Guillaume> medoc.inria.fr for /home/yquem/.directory (/): no export entry

Guillaume> I guess they are caused by graphical file manager trying to find
Guillaume> metadata directories when displaying mount points. Is there a way to
Guillaume> prevent them ?

No, there is not a way to prevent them.  The basic problem is that the
requests could be valid;  how is the automounter to know which
directory lookups to ignore?  haldaemon has been responsible for some
problems in the past, but its authors chose to include special cases
for automount-managed directories.  This is not the most ideal
solution, I guess, but it is the most attractive from the autofs point
of view.

Can you find out which applications are causing these lookups and file
bugs in their respective bug trackers?

-Jeff

Re: excluding some pattern to prevent useless mount request

[Guillaume Rousse]

Jeff Moyer wrote:
> ==> On Mon, 27 Nov 2006 17:40:54 +0100, Guillaume Rousse <Guillaume.Rousse@inria.fr> said:
> 
> Guillaume> While looking in server's log, I found many such mount requests:
> Guillaume> Nov 26 04:23:19 yquem rpc.mountd: refused mount request from
> Guillaume> chouchen.inria.fr for /home/yquem/.hidden (/): no export entry
> Guillaume> ...
> Guillaume> Nov 26 04:22:33 yquem rpc.mountd: refused mount request from
> Guillaume> medoc.inria.fr for /home/yquem/.directory (/): no export entry
> 
> Guillaume> I guess they are caused by graphical file manager trying to find
> Guillaume> metadata directories when displaying mount points. Is there a way to
> Guillaume> prevent them ?
> 
> No, there is not a way to prevent them.  The basic problem is that the
> requests could be valid;  how is the automounter to know which
> directory lookups to ignore? 
If it could be configured locally, the automounter would know :)

> haldaemon has been responsible for some
> problems in the past, but its authors chose to include special cases
> for automount-managed directories.  This is not the most ideal
> solution, I guess, but it is the most attractive from the autofs point
> of view.
> 
> Can you find out which applications are causing these lookups and file
> bugs in their respective bug trackers?
I don't feel very confident trying to convince konqueror developer they
have to always test first if the directory they are going to probe is
automountable... This seems to be both a exception, and something more
easily implementable in a centralized way in automounter.