2012/11/8 J. Bruce Fields <bfields@xxxxxxxxxxxx>: > On Thu, Nov 08, 2012 at 11:03:22AM +0800, Jack Wang wrote: >> 2012/11/8 J. Bruce Fields <bfields@xxxxxxxxxxxx>: >> > On Wed, Nov 07, 2012 at 09:05:51AM +0800, Jack Wang wrote: >> >> 2012/11/7 Martín Cigorraga <msx@xxxxxxxxxxxx>: >> >> > On Mon, Nov 5, 2012 at 10:27 PM, Jack Wang <jack.wang.usish@xxxxxxxxx> >> >> > wrote: >> >> >> >> >> >> Hi all, >> >> >> >> >> >> Anyone who can kindly give some suggestion? or we just put question to >> >> >> the wrong list? >> >> >> >> >> >> Thanks. >> >> >> >> >> >> Jack >> >> >> >> >> >> 2012/11/1 dahai_tian <dahai_tian@xxxxxxxxx>: >> >> >> > Hi all: >> >> >> > I mount a local directory to a nfs server. When nfs server is >> >> >> > stopped >> >> >> > for some accidental cause, I try to umount the mount point, the umount >> >> >> > command will hang and 'time out' messages are continually printed in the >> >> >> > terminal. How can I avoid hanging in this case? BTW, This issue does not >> >> >> > exist when I changed nfs version from 4 to 3. >> >> >> > Look forward to your response, thanks! >> >> >> >> > >> >> > Hi Jack, use the -l (lazy) switch: >> >> > # umount -l {your mounted nfs share} >> >> > >> >> > Also, I wouldn't recommend 'just to kill' the NFS proces(ses) as it (them) >> >> > may leave shared memory a mess. >> >> > (And I would like to see how do you make to mount any NFS share again >> >> > without booting) >> >> >> >> Thanks Martin for kindly help. >> >> >> >> umount -l did work, but why nfsv3 do not have this problem? >> > >> > I don't know what it is exactly. But in general I wouldn't expect >> > umount to work when the server's unavailable. >> > >> > --b. >> >> Thanks for your time Bruce, but there are chance the server >> unavailable , why you think umount don't work is expected? > > It's tough for the client to deal safely with an unreachable server. > Worst case it may still have data in its buffers that applications have > written but that hasn't yet made it to the server. > > If you don't have any opens or other state on the filesystem any more, > there might be things that could be done to make it easier to umount in > that case--I haven't thought it through--but I doubt that's a high > priority, so it's likely up to you to write the code and persuade people > it's worthwhile. > > --b. Thanks for explaination. Jack -- 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