On Tue, 2009-03-31 at 17:16 -0700, Linus Torvalds wrote: > > On Tue, 31 Mar 2009, Trond Myklebust wrote: > > > > However, if your mount point filehandle has expired, then chances are > > that you'll probably have to look up all the filehandles in the mount > > path again too, so why should you cache all that information? > > No, that' not the caching I was thinking about. > > I meant literally just the "namespace" thing. Keeping that around as long > as you have a export active. How many of those do you have on your average > nfs server? Why not just create those nfsd namespaces once for each > export at startup, and destroy them at exit. Wait, I'm confused now. None of this is running on the server. The private namespace is being created on the NFSv4 client, which only needs to resolve a mount path as part of a sys_mount() call. Unlike NFSv2/v3, there is no MOUNT protocol to translate an NFSv4 mount path into a filehandle. Instead the client has an operation to fetch the 'root filehandle' (i.e. the filehandle representing 'mount -t nfs4 foo:/') and then does a series of LOOKUP rpc calls to resolve the rest of the path. Cheers Trond -- Trond Myklebust Linux NFS client maintainer NetApp Trond.Myklebust@xxxxxxxxxx www.netapp.com -- To unsubscribe from this list: send the line "unsubscribe linux-fsdevel" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
