On Mon, 2013-07-08 at 09:58 +1000, NeilBrown wrote:
>
> This patch adds a "nosharetransport" option to allow two different
> mounts from the same server to use different transports.
> If the mounts use NFSv4, or are of the same filesystem, then
> "nosharecache" must be used as well.
Won't this interfere with the recently added NFSv4 trunking detection?
Also, how will it work with NFSv4.1 sessions? The server will usually
require a BIND_CONN_TO_SESSION when new TCP connections attempt to
attach to an existing session.
> There are at least two circumstances where it might be desirable
> to use separate transports:
>
> 1/ If the NFS server can get into a state where it will ignore
> requests for one filesystem while servicing request for another,
> then using separate connections for the separate filesystems can
> stop problems with one affecting access to the other.
>
> This is particularly relevant for NetApp filers where one filesystem
> has been "suspended". Requests to that filesystem will be dropped
> (rather than the more correct NFS3ERR_JUKEBOX). This currently
> interferes with other filesystems.
This is a known issue that really needs to be fixed on the server, not
on the client. As far as I know, work is already underway to fix this.
> 2/ If a very fast network is used with a many-processor client, a
> single TCP connection can present a bottle neck which reduces total
> throughput. Using multiple TCP connections (one per mount) removes
> the bottleneck.
> An alternate workaround is to configure multiple virtual IP
> addresses on the server and mount each filesystem from a different
> IP. This is effective (throughput goes up) but an unnecessary
> administrative burden.
As I understand it, using multiple simultaneous TCP connections between
the same endpoints also adds a risk that the congestion windows will
interfere. Do you have numbers to back up the claim of a performance
improvement?
The other issue I can think of is that for NFS versions < 4.1, this may
cause the server to allocate more resources per client in the form of
replay caches etc.
> Signed-off-by: NeilBrown <neilb@xxxxxxx>
>
> ---
> Is this a good idea? Bad idea? Have I missed something important?
>
> NeilBrown
>
>
> diff --git a/fs/nfs/client.c b/fs/nfs/client.c
> index c513b0c..64e3f39 100644
> --- a/fs/nfs/client.c
> +++ b/fs/nfs/client.c
> @@ -403,8 +403,13 @@ static struct nfs_client *nfs_match_client(const struct nfs_client_initdata *dat
> const struct sockaddr *sap = data->addr;
> struct nfs_net *nn = net_generic(data->net, nfs_net_id);
>
> + if (test_bit(NFS_CS_NO_SHARE, &data->init_flags))
> + return NULL;
> +
> list_for_each_entry(clp, &nn->nfs_client_list, cl_share_link) {
> const struct sockaddr *clap = (struct sockaddr *)&clp->cl_addr;
> + if (test_bit(NFS_CS_NO_SHARE,&clp->cl_flags))
> + continue;
> /* Don't match clients that failed to initialise properly */
> if (clp->cl_cons_state < 0)
> continue;
> @@ -753,6 +758,8 @@ static int nfs_init_server(struct nfs_server *server,
> data->timeo, data->retrans);
> if (data->flags & NFS_MOUNT_NORESVPORT)
> set_bit(NFS_CS_NORESVPORT, &cl_init.init_flags);
> + if (data->flags & NFS_MOUNT_NOSHARE_XPRT)
> + set_bit(NFS_CS_NO_SHARE, &cl_init.init_flags);
> if (server->options & NFS_OPTION_MIGRATION)
> set_bit(NFS_CS_MIGRATION, &cl_init.init_flags);
>
> diff --git a/fs/nfs/super.c b/fs/nfs/super.c
> index 2d7525f..d9141d8 100644
> --- a/fs/nfs/super.c
> +++ b/fs/nfs/super.c
> @@ -88,6 +88,7 @@ enum {
> Opt_acl, Opt_noacl,
> Opt_rdirplus, Opt_nordirplus,
> Opt_sharecache, Opt_nosharecache,
> + Opt_sharetransport, Opt_nosharetransport,
> Opt_resvport, Opt_noresvport,
> Opt_fscache, Opt_nofscache,
> Opt_migration, Opt_nomigration,
> @@ -146,6 +147,8 @@ static const match_table_t nfs_mount_option_tokens = {
> { Opt_nordirplus, "nordirplus" },
> { Opt_sharecache, "sharecache" },
> { Opt_nosharecache, "nosharecache" },
> + { Opt_sharetransport, "sharetransport"},
> + { Opt_nosharetransport, "nosharetransport"},
> { Opt_resvport, "resvport" },
> { Opt_noresvport, "noresvport" },
> { Opt_fscache, "fsc" },
> @@ -634,6 +637,7 @@ static void nfs_show_mount_options(struct seq_file *m, struct nfs_server *nfss,
> { NFS_MOUNT_NOACL, ",noacl", "" },
> { NFS_MOUNT_NORDIRPLUS, ",nordirplus", "" },
> { NFS_MOUNT_UNSHARED, ",nosharecache", "" },
> + { NFS_MOUNT_NOSHARE_XPRT, ",nosharetransport", ""},
> { NFS_MOUNT_NORESVPORT, ",noresvport", "" },
> { 0, NULL, NULL }
> };
> @@ -1239,6 +1243,12 @@ static int nfs_parse_mount_options(char *raw,
> case Opt_nosharecache:
> mnt->flags |= NFS_MOUNT_UNSHARED;
> break;
> + case Opt_sharetransport:
> + mnt->flags &= ~NFS_MOUNT_NOSHARE_XPRT;
> + break;
> + case Opt_nosharetransport:
> + mnt->flags |= NFS_MOUNT_NOSHARE_XPRT;
> + break;
> case Opt_resvport:
> mnt->flags &= ~NFS_MOUNT_NORESVPORT;
> break;
> diff --git a/include/linux/nfs_fs_sb.h b/include/linux/nfs_fs_sb.h
> index 3b7fa2a..9e9d7d3 100644
> --- a/include/linux/nfs_fs_sb.h
> +++ b/include/linux/nfs_fs_sb.h
> @@ -41,6 +41,7 @@ struct nfs_client {
> #define NFS_CS_DISCRTRY 1 /* - disconnect on RPC retry */
> #define NFS_CS_MIGRATION 2 /* - transparent state migr */
> #define NFS_CS_INFINITE_SLOTS 3 /* - don't limit TCP slots */
> +#define NFS_CS_NO_SHARE 4 /* - don't share across mounts */
> struct sockaddr_storage cl_addr; /* server identifier */
> size_t cl_addrlen;
> char * cl_hostname; /* hostname of server */
> diff --git a/include/uapi/linux/nfs_mount.h b/include/uapi/linux/nfs_mount.h
> index 576bddd..81c49ff 100644
> --- a/include/uapi/linux/nfs_mount.h
> +++ b/include/uapi/linux/nfs_mount.h
> @@ -73,5 +73,6 @@ struct nfs_mount_data {
>
> #define NFS_MOUNT_LOCAL_FLOCK 0x100000
> #define NFS_MOUNT_LOCAL_FCNTL 0x200000
> +#define NFS_MOUNT_NOSHARE_XPRT 0x400000
>
> #endif
--
Trond Myklebust
Linux NFS client maintainer
NetApp
Trond.Myklebust@xxxxxxxxxx
www.netapp.com
��.n��������+%������w��{.n�����{��w��� jg��� �����ݢj����G���� ���j:+v���w�m������w�������h�����٥
