[PATCH v3 0/6] do not collapse trunkable transports

[PATCH v3 0/6] do not collapse trunkable transports

[Olga Kornievskaia]

From: Olga Kornievskaia <kolga@netapp.com>

This patch series attempts to allow for new mounts that are to the
same server (ie nfsv4.1+ session trunkable servers) but different
network addresses to use connections associated with those mounts
but still use the same client structure.

A new mount options, "max_connect", controls how many extra transports
can be added to an existing client, with maximum of 16 such transports.

v3:
-- add a new counter to xprt_switch to keep track of transports with
unique addresses
-- control of enforcing the limit is moved into the sunrpc layer
into the function rpc_clnt_test_and_add_xprt
-- after a trunking transport is created if mount request asked for
nconnect connections, create as many as the original/first mount had

Olga Kornievskaia (6):
  SUNRPC keep track of number of transports to unique addresses
  SUNRPC add xps_nunique_destaddr_xprts to xprt_switch_info in sysfs
  NFSv4 introduce max_connect mount options
  SUNRPC enforce creation of no more than max_connect xprts
  NFSv4 add network transport when session trunking is detected
  NFSv4 allow for nconnect value of trunkable transport

 fs/nfs/client.c                      |  2 ++
 fs/nfs/fs_context.c                  |  7 ++++
 fs/nfs/internal.h                    |  2 ++
 fs/nfs/nfs4client.c                  | 50 ++++++++++++++++++++++++++--
 fs/nfs/super.c                       |  2 ++
 include/linux/nfs_fs.h               |  5 +++
 include/linux/nfs_fs_sb.h            |  1 +
 include/linux/sunrpc/clnt.h          |  2 ++
 include/linux/sunrpc/xprtmultipath.h |  1 +
 net/sunrpc/clnt.c                    | 11 +++++-
 net/sunrpc/sysfs.c                   |  4 ++-
 net/sunrpc/xprtmultipath.c           |  1 +
 12 files changed, 84 insertions(+), 4 deletions(-)

-- 
2.27.0

[PATCH v3 1/6] SUNRPC keep track of number of transports to unique addresses

[Olga Kornievskaia]

From: Olga Kornievskaia <kolga@netapp.com>

Currently, xprt_switch keeps a number of all xprts (xps_nxprts)
that were added to the switch regardless of whethere it's an
nconnect transport or a transport to a trunkable address.
Introduce a new counter to keep track of transports to unique
destination addresses per xprt_switch.

Signed-off-by: Olga Kornievskaia <kolga@netapp.com>
---
 include/linux/sunrpc/xprtmultipath.h | 1 +
 net/sunrpc/clnt.c                    | 2 +-
 net/sunrpc/xprtmultipath.c           | 1 +
 3 files changed, 3 insertions(+), 1 deletion(-)

diff --git a/include/linux/sunrpc/xprtmultipath.h b/include/linux/sunrpc/xprtmultipath.h
index b19addc8b715..bbb8a5fa0816 100644
--- a/include/linux/sunrpc/xprtmultipath.h
+++ b/include/linux/sunrpc/xprtmultipath.h
@@ -18,6 +18,7 @@ struct rpc_xprt_switch {
 	unsigned int		xps_id;
 	unsigned int		xps_nxprts;
 	unsigned int		xps_nactive;
+	unsigned int		xps_nunique_destaddr_xprts;
 	atomic_long_t		xps_queuelen;
 	struct list_head	xps_xprt_list;
 
diff --git a/net/sunrpc/clnt.c b/net/sunrpc/clnt.c
index 9bf820bad84c..e6801a481d02 100644
--- a/net/sunrpc/clnt.c
+++ b/net/sunrpc/clnt.c
@@ -2761,7 +2761,7 @@ int rpc_clnt_test_and_add_xprt(struct rpc_clnt *clnt,
 
 	task = rpc_call_null_helper(clnt, xprt, NULL, RPC_TASK_ASYNC,
 			&rpc_cb_add_xprt_call_ops, data);
-
+	data->xps->xps_nunique_destaddr_xprts++;
 	rpc_put_task(task);
 success:
 	return 1;
diff --git a/net/sunrpc/xprtmultipath.c b/net/sunrpc/xprtmultipath.c
index 07e76ae1028a..584349c8cad4 100644
--- a/net/sunrpc/xprtmultipath.c
+++ b/net/sunrpc/xprtmultipath.c
@@ -138,6 +138,7 @@ struct rpc_xprt_switch *xprt_switch_alloc(struct rpc_xprt *xprt,
 		xps->xps_iter_ops = &rpc_xprt_iter_singular;
 		rpc_sysfs_xprt_switch_setup(xps, xprt, gfp_flags);
 		xprt_switch_add_xprt_locked(xps, xprt);
+		xps->xps_nunique_destaddr_xprts = 1;
 		rpc_sysfs_xprt_setup(xps, xprt, gfp_flags);
 	}
 
-- 
2.27.0

[PATCH v3 2/6] SUNRPC add xps_nunique_destaddr_xprts to xprt_switch_info in sysfs

[Olga Kornievskaia]

From: Olga Kornievskaia <kolga@netapp.com>

In sysfs's xprt_switch_info attribute also display the value of
number of transports with unique destination addresses for this
xprt_switch.

Signed-off-by: Olga Kornievskaia <kolga@netapp.com>
---
 net/sunrpc/sysfs.c | 4 +++-
 1 file changed, 3 insertions(+), 1 deletion(-)

diff --git a/net/sunrpc/sysfs.c b/net/sunrpc/sysfs.c
index 2fbaba27d5c6..45c0cf18360d 100644
--- a/net/sunrpc/sysfs.c
+++ b/net/sunrpc/sysfs.c
@@ -164,8 +164,10 @@ static ssize_t rpc_sysfs_xprt_switch_info_show(struct kobject *kobj,
 
 	if (!xprt_switch)
 		return 0;
-	ret = sprintf(buf, "num_xprts=%u\nnum_active=%u\nqueue_len=%ld\n",
+	ret = sprintf(buf, "num_xprts=%u\nnum_active=%u\n"
+		      "num_unique_destaddr=%u\nqueue_len=%ld\n",
 		      xprt_switch->xps_nxprts, xprt_switch->xps_nactive,
+		      xprt_switch->xps_nunique_destaddr_xprts,
 		      atomic_long_read(&xprt_switch->xps_queuelen));
 	xprt_switch_put(xprt_switch);
 	return ret + 1;
-- 
2.27.0

[PATCH v3 3/6] NFSv4 introduce max_connect mount options

[Olga Kornievskaia]

From: Olga Kornievskaia <kolga@netapp.com>

This option will control up to how many xprts can the client
establish to the server with a distinct address (that means
nconnect connections are not counted towards this new limit).
This patch is setting up nfs structures to keeep track of the
max_connect limit (does not enforce it).

The default value is kept at 1 so that no current mounts that
don't want any additional connections would be effected. The
maximum value is set at 16.

Mounts to DS are not limited to default value of 1 but instead
set to the maximum default value of 16 (NFS_MAX_TRANSPORTS).

Signed-off-by: Olga Kornievskaia <kolga@netapp.com>
---
 fs/nfs/client.c           |  1 +
 fs/nfs/fs_context.c       |  7 +++++++
 fs/nfs/internal.h         |  2 ++
 fs/nfs/nfs4client.c       | 12 ++++++++++--
 fs/nfs/super.c            |  2 ++
 include/linux/nfs_fs.h    |  5 +++++
 include/linux/nfs_fs_sb.h |  1 +
 7 files changed, 28 insertions(+), 2 deletions(-)

diff --git a/fs/nfs/client.c b/fs/nfs/client.c
index 330f65727c45..486dec59972b 100644
--- a/fs/nfs/client.c
+++ b/fs/nfs/client.c
@@ -179,6 +179,7 @@ struct nfs_client *nfs_alloc_client(const struct nfs_client_initdata *cl_init)
 
 	clp->cl_proto = cl_init->proto;
 	clp->cl_nconnect = cl_init->nconnect;
+	clp->cl_max_connect = cl_init->max_connect ? cl_init->max_connect : 1;
 	clp->cl_net = get_net(cl_init->net);
 
 	clp->cl_principal = "*";
diff --git a/fs/nfs/fs_context.c b/fs/nfs/fs_context.c
index d95c9a39bc70..0d444a90f513 100644
--- a/fs/nfs/fs_context.c
+++ b/fs/nfs/fs_context.c
@@ -60,6 +60,7 @@ enum nfs_param {
 	Opt_mountvers,
 	Opt_namelen,
 	Opt_nconnect,
+	Opt_max_connect,
 	Opt_port,
 	Opt_posix,
 	Opt_proto,
@@ -158,6 +159,7 @@ static const struct fs_parameter_spec nfs_fs_parameters[] = {
 	fsparam_u32   ("mountvers",	Opt_mountvers),
 	fsparam_u32   ("namlen",	Opt_namelen),
 	fsparam_u32   ("nconnect",	Opt_nconnect),
+	fsparam_u32   ("max_connect",	Opt_max_connect),
 	fsparam_string("nfsvers",	Opt_vers),
 	fsparam_u32   ("port",		Opt_port),
 	fsparam_flag_no("posix",	Opt_posix),
@@ -770,6 +772,11 @@ static int nfs_fs_context_parse_param(struct fs_context *fc,
 			goto out_of_bounds;
 		ctx->nfs_server.nconnect = result.uint_32;
 		break;
+	case Opt_max_connect:
+		if (result.uint_32 < 1 || result.uint_32 > NFS_MAX_TRANSPORTS)
+			goto out_of_bounds;
+		ctx->nfs_server.max_connect = result.uint_32;
+		break;
 	case Opt_lookupcache:
 		switch (result.uint_32) {
 		case Opt_lookupcache_all:
diff --git a/fs/nfs/internal.h b/fs/nfs/internal.h
index a36af04188c2..66fc936834f2 100644
--- a/fs/nfs/internal.h
+++ b/fs/nfs/internal.h
@@ -67,6 +67,7 @@ struct nfs_client_initdata {
 	int proto;
 	u32 minorversion;
 	unsigned int nconnect;
+	unsigned int max_connect;
 	struct net *net;
 	const struct rpc_timeout *timeparms;
 	const struct cred *cred;
@@ -121,6 +122,7 @@ struct nfs_fs_context {
 		int			port;
 		unsigned short		protocol;
 		unsigned short		nconnect;
+		unsigned short		max_connect;
 		unsigned short		export_path_len;
 	} nfs_server;
 
diff --git a/fs/nfs/nfs4client.c b/fs/nfs/nfs4client.c
index 28431acd1230..270caa1805a2 100644
--- a/fs/nfs/nfs4client.c
+++ b/fs/nfs/nfs4client.c
@@ -865,6 +865,7 @@ static int nfs4_set_client(struct nfs_server *server,
 		const char *ip_addr,
 		int proto, const struct rpc_timeout *timeparms,
 		u32 minorversion, unsigned int nconnect,
+		unsigned int max_connect,
 		struct net *net)
 {
 	struct nfs_client_initdata cl_init = {
@@ -883,6 +884,8 @@ static int nfs4_set_client(struct nfs_server *server,
 
 	if (minorversion == 0)
 		__set_bit(NFS_CS_REUSEPORT, &cl_init.init_flags);
+	else
+		cl_init.max_connect = max_connect;
 	if (proto == XPRT_TRANSPORT_TCP)
 		cl_init.nconnect = nconnect;
 
@@ -952,8 +955,10 @@ struct nfs_client *nfs4_set_ds_client(struct nfs_server *mds_srv,
 		return ERR_PTR(-EINVAL);
 	cl_init.hostname = buf;
 
-	if (mds_clp->cl_nconnect > 1 && ds_proto == XPRT_TRANSPORT_TCP)
+	if (mds_clp->cl_nconnect > 1 && ds_proto == XPRT_TRANSPORT_TCP) {
 		cl_init.nconnect = mds_clp->cl_nconnect;
+		cl_init.max_connect = NFS_MAX_TRANSPORTS;
+	}
 
 	if (mds_srv->flags & NFS_MOUNT_NORESVPORT)
 		__set_bit(NFS_CS_NORESVPORT, &cl_init.init_flags);
@@ -1122,6 +1127,7 @@ static int nfs4_init_server(struct nfs_server *server, struct fs_context *fc)
 				&timeparms,
 				ctx->minorversion,
 				ctx->nfs_server.nconnect,
+				ctx->nfs_server.max_connect,
 				fc->net_ns);
 	if (error < 0)
 		return error;
@@ -1211,6 +1217,7 @@ struct nfs_server *nfs4_create_referral_server(struct fs_context *fc)
 				parent_server->client->cl_timeout,
 				parent_client->cl_mvops->minor_version,
 				parent_client->cl_nconnect,
+				parent_client->cl_max_connect,
 				parent_client->cl_net);
 	if (!error)
 		goto init_server;
@@ -1226,6 +1233,7 @@ struct nfs_server *nfs4_create_referral_server(struct fs_context *fc)
 				parent_server->client->cl_timeout,
 				parent_client->cl_mvops->minor_version,
 				parent_client->cl_nconnect,
+				parent_client->cl_max_connect,
 				parent_client->cl_net);
 	if (error < 0)
 		goto error;
@@ -1323,7 +1331,7 @@ int nfs4_update_server(struct nfs_server *server, const char *hostname,
 	error = nfs4_set_client(server, hostname, sap, salen, buf,
 				clp->cl_proto, clnt->cl_timeout,
 				clp->cl_minorversion,
-				clp->cl_nconnect, net);
+				clp->cl_nconnect, clp->cl_max_connect, net);
 	clear_bit(NFS_MIG_TSM_POSSIBLE, &server->mig_status);
 	if (error != 0) {
 		nfs_server_insert_lists(server);
diff --git a/fs/nfs/super.c b/fs/nfs/super.c
index fe58525cfed4..e65c83494c05 100644
--- a/fs/nfs/super.c
+++ b/fs/nfs/super.c
@@ -480,6 +480,8 @@ static void nfs_show_mount_options(struct seq_file *m, struct nfs_server *nfss,
 	if (clp->cl_nconnect > 0)
 		seq_printf(m, ",nconnect=%u", clp->cl_nconnect);
 	if (version == 4) {
+		if (clp->cl_max_connect > 1)
+			seq_printf(m, ",max_connect=%u", clp->cl_max_connect);
 		if (nfss->port != NFS_PORT)
 			seq_printf(m, ",port=%u", nfss->port);
 	} else
diff --git a/include/linux/nfs_fs.h b/include/linux/nfs_fs.h
index ffba254d2098..c13414a8dfde 100644
--- a/include/linux/nfs_fs.h
+++ b/include/linux/nfs_fs.h
@@ -40,6 +40,11 @@
 
 #include <linux/mempool.h>
 
+/*
+ * These are the default for number of transports to different server IPs
+ */
+#define NFS_MAX_TRANSPORTS 16
+
 /*
  * These are the default flags for swap requests
  */
diff --git a/include/linux/nfs_fs_sb.h b/include/linux/nfs_fs_sb.h
index d71a0e90faeb..2a9acbfe00f0 100644
--- a/include/linux/nfs_fs_sb.h
+++ b/include/linux/nfs_fs_sb.h
@@ -62,6 +62,7 @@ struct nfs_client {
 
 	u32			cl_minorversion;/* NFSv4 minorversion */
 	unsigned int		cl_nconnect;	/* Number of connections */
+	unsigned int		cl_max_connect; /* max number of xprts allowed */
 	const char *		cl_principal;  /* used for machine cred */
 
 #if IS_ENABLED(CONFIG_NFS_V4)
-- 
2.27.0

[PATCH v3 4/6] SUNRPC enforce creation of no more than max_connect xprts

[Olga Kornievskaia]

From: Olga Kornievskaia <kolga@netapp.com>

If we are adding new transports via rpc_clnt_test_and_add_xprt()
then check if we've reached the limit. Currently only pnfs path
adds transports via that function but this is done in
preparation when the client would add new transports when
session trunking is detected. A warning is logged if the
limit is reached.

Signed-off-by: Olga Kornievskaia <kolga@netapp.com>
---
 fs/nfs/client.c             | 1 +
 include/linux/sunrpc/clnt.h | 2 ++
 net/sunrpc/clnt.c           | 9 +++++++++
 3 files changed, 12 insertions(+)

diff --git a/fs/nfs/client.c b/fs/nfs/client.c
index 486dec59972b..23e165d5ec9c 100644
--- a/fs/nfs/client.c
+++ b/fs/nfs/client.c
@@ -541,6 +541,7 @@ int nfs_create_rpc_client(struct nfs_client *clp,
 
 	clnt->cl_principal = clp->cl_principal;
 	clp->cl_rpcclient = clnt;
+	clnt->cl_max_connect = clp->cl_max_connect;
 	return 0;
 }
 EXPORT_SYMBOL_GPL(nfs_create_rpc_client);
diff --git a/include/linux/sunrpc/clnt.h b/include/linux/sunrpc/clnt.h
index 8b5d5c97553e..0bbf873b225a 100644
--- a/include/linux/sunrpc/clnt.h
+++ b/include/linux/sunrpc/clnt.h
@@ -81,6 +81,7 @@ struct rpc_clnt {
 		struct work_struct	cl_work;
 	};
 	const struct cred	*cl_cred;
+	unsigned int		cl_max_connect; /* max number of transports not to the same IP */
 };
 
 /*
@@ -135,6 +136,7 @@ struct rpc_create_args {
 	char			*client_name;
 	struct svc_xprt		*bc_xprt;	/* NFSv4.1 backchannel */
 	const struct cred	*cred;
+	unsigned int		max_connect;
 };
 
 struct rpc_add_xprt_test {
diff --git a/net/sunrpc/clnt.c b/net/sunrpc/clnt.c
index e6801a481d02..4d8fd9d9c264 100644
--- a/net/sunrpc/clnt.c
+++ b/net/sunrpc/clnt.c
@@ -2749,6 +2749,15 @@ int rpc_clnt_test_and_add_xprt(struct rpc_clnt *clnt,
 	struct rpc_cb_add_xprt_calldata *data;
 	struct rpc_task *task;
 
+	if (xps->xps_nunique_destaddr_xprts + 1 > clnt->cl_max_connect) {
+		rcu_read_lock();
+		pr_warn("SUNRPC: reached max allowed number (%d) did not add "
+			"transport to server: %s\n", clnt->cl_max_connect,
+			rcu_dereference(xprt->address_strings[RPC_DISPLAY_ADDR]));
+		rcu_read_unlock();
+		return -EINVAL;
+	}
+
 	data = kmalloc(sizeof(*data), GFP_NOFS);
 	if (!data)
 		return -ENOMEM;
-- 
2.27.0

[PATCH v3 5/6] NFSv4 add network transport when session trunking is detected

[Olga Kornievskaia]

From: Olga Kornievskaia <kolga@netapp.com>

After trunking is discovered in nfs4_discover_server_trunking(),
add the transport to the old client structure if the allowed limit
of transports has not been reached.

An example: there exists a multi-homed server and client mounts
one server address and some volume and then doest another mount to
a different address of the same server and perhaps a different
volume. Previously, the client checks that this is a session
trunkable servers (same server), and removes the newly created
client structure along with its transport. Now, the client
adds the connection from the 2nd mount into the xprt switch of
the existing client (it leads to having 2 available connections).

Signed-off-by: Olga Kornievskaia <kolga@netapp.com>
---
 fs/nfs/nfs4client.c | 29 +++++++++++++++++++++++++++++
 1 file changed, 29 insertions(+)

diff --git a/fs/nfs/nfs4client.c b/fs/nfs/nfs4client.c
index 270caa1805a2..af57332503be 100644
--- a/fs/nfs/nfs4client.c
+++ b/fs/nfs/nfs4client.c
@@ -402,6 +402,33 @@ static int nfs4_init_client_minor_version(struct nfs_client *clp)
 	return nfs4_init_callback(clp);
 }
 
+static void nfs4_add_trunk(struct nfs_client *clp, struct nfs_client *old)
+{
+	struct sockaddr_storage clp_addr, old_addr;
+	struct sockaddr *clp_sap = (struct sockaddr *)&clp_addr;
+	struct sockaddr *old_sap = (struct sockaddr *)&old_addr;
+	size_t clp_salen;
+	struct xprt_create xprt_args = {
+		.ident = old->cl_proto,
+		.net = old->cl_net,
+		.servername = old->cl_hostname,
+	};
+
+	if (clp->cl_proto != old->cl_proto)
+		return;
+	clp_salen = rpc_peeraddr(clp->cl_rpcclient, clp_sap, sizeof(clp_addr));
+	rpc_peeraddr(old->cl_rpcclient, old_sap, sizeof(old_addr));
+
+	if (clp_addr.ss_family != old_addr.ss_family)
+		return;
+
+	xprt_args.dstaddr = clp_sap;
+	xprt_args.addrlen = clp_salen;
+
+	rpc_clnt_add_xprt(old->cl_rpcclient, &xprt_args,
+			  rpc_clnt_test_and_add_xprt, NULL);
+}
+
 /**
  * nfs4_init_client - Initialise an NFS4 client record
  *
@@ -436,6 +463,8 @@ struct nfs_client *nfs4_init_client(struct nfs_client *clp,
 		 * won't try to use it.
 		 */
 		nfs_mark_client_ready(clp, -EPERM);
+		if (old->cl_mvops->session_trunk)
+			nfs4_add_trunk(clp, old);
 	}
 	clear_bit(NFS_CS_TSM_POSSIBLE, &clp->cl_flags);
 	nfs_put_client(clp);
-- 
2.27.0

[PATCH v3 6/6] NFSv4 allow for nconnect value of trunkable transport

[Olga Kornievskaia]

From: Olga Kornievskaia <kolga@netapp.com>

If the new mount asked for nconnect mount, then create old client
number of connections to the destination address that has been
established as the same server with trunkable address.

Signed-off-by: Olga Kornievskaia <kolga@netapp.com>

--- There might be a number of objection to this patch. One
I can think of is that this patch creates the nconnects based on
whether or not the new mount asked for nconnect instead of
unconditionally creating nconnect number of connections. The patch
still creates nconnect connections based on the original value
instead of picking the value of clp->cl_nconnect. I would have
preferred that would be done. I don't see what can be wrong with
using the new value. But I feared to go against what was objected
before. My preference would be to (1) create clp->cl_nconnect
connections or (2) not use this patch at all or (3) use as is
here (meaning at least not create extra connections unless asked
for by the mount).
---
 fs/nfs/nfs4client.c | 9 +++++++++
 1 file changed, 9 insertions(+)

diff --git a/fs/nfs/nfs4client.c b/fs/nfs/nfs4client.c
index af57332503be..50fa9d53b444 100644
--- a/fs/nfs/nfs4client.c
+++ b/fs/nfs/nfs4client.c
@@ -427,6 +427,15 @@ static void nfs4_add_trunk(struct nfs_client *clp, struct nfs_client *old)
 
 	rpc_clnt_add_xprt(old->cl_rpcclient, &xprt_args,
 			  rpc_clnt_test_and_add_xprt, NULL);
+
+	if (clp->cl_nconnect > 1) {
+		int i;
+
+		for (i = 0; i < old->cl_nconnect - 1; i++)
+			if (rpc_clnt_add_xprt(old->cl_rpcclient, &xprt_args,
+					      NULL, NULL) < 0)
+				break;
+	}
 }
 
 /**
-- 
2.27.0

Re: [PATCH v3 0/6] do not collapse trunkable transports

[Olga Kornievskaia]

On Tue, Jun 15, 2021 at 9:10 PM Olga Kornievskaia
<olga.kornievskaia@gmail.com> wrote:
>
> From: Olga Kornievskaia <kolga@netapp.com>
>
> This patch series attempts to allow for new mounts that are to the
> same server (ie nfsv4.1+ session trunkable servers) but different
> network addresses to use connections associated with those mounts
> but still use the same client structure.
>
> A new mount options, "max_connect", controls how many extra transports
> can be added to an existing client, with maximum of 16 such transports.
>
> v3:
> -- add a new counter to xprt_switch to keep track of transports with
> unique addresses
> -- control of enforcing the limit is moved into the sunrpc layer
> into the function rpc_clnt_test_and_add_xprt
> -- after a trunking transport is created if mount request asked for
> nconnect connections, create as many as the original/first mount had

Any comments on/objections to this version?

>
> Olga Kornievskaia (6):
>   SUNRPC keep track of number of transports to unique addresses
>   SUNRPC add xps_nunique_destaddr_xprts to xprt_switch_info in sysfs
>   NFSv4 introduce max_connect mount options
>   SUNRPC enforce creation of no more than max_connect xprts
>   NFSv4 add network transport when session trunking is detected
>   NFSv4 allow for nconnect value of trunkable transport
>
>  fs/nfs/client.c                      |  2 ++
>  fs/nfs/fs_context.c                  |  7 ++++
>  fs/nfs/internal.h                    |  2 ++
>  fs/nfs/nfs4client.c                  | 50 ++++++++++++++++++++++++++--
>  fs/nfs/super.c                       |  2 ++
>  include/linux/nfs_fs.h               |  5 +++
>  include/linux/nfs_fs_sb.h            |  1 +
>  include/linux/sunrpc/clnt.h          |  2 ++
>  include/linux/sunrpc/xprtmultipath.h |  1 +
>  net/sunrpc/clnt.c                    | 11 +++++-
>  net/sunrpc/sysfs.c                   |  4 ++-
>  net/sunrpc/xprtmultipath.c           |  1 +
>  12 files changed, 84 insertions(+), 4 deletions(-)
>
> --
> 2.27.0
>

Re: [PATCH v3 6/6] NFSv4 allow for nconnect value of trunkable transport

[Olga Kornievskaia]

Hello folks (Chuck, Trond have graciously participated previously and
I look forward to hearing back from you),

I would like to restart the conversation about the client's behaviour
with regards to discovered truncable transports.

I have put forth a version and patch 6 is the one that deals with the
combination of trunkable transport and nconnect and I look for
feedback from the community. I have written my views in the comments
in the patch.

My current position is: given that it's unclear how to reconcile
managing connections to the same IP together with connections to other
IPs. I propose we don't bring in nconnect  in play when truncable
transport is discovered (basically option #2 in my comments below). I
believe that's the simplest thing to do now and we can add nconnect
behaviour later.

Thank you.


On Tue, Jun 15, 2021 at 9:10 PM Olga Kornievskaia
<olga.kornievskaia@gmail.com> wrote:
>
> From: Olga Kornievskaia <kolga@netapp.com>
>
> If the new mount asked for nconnect mount, then create old client
> number of connections to the destination address that has been
> established as the same server with trunkable address.
>
> Signed-off-by: Olga Kornievskaia <kolga@netapp.com>
>
> --- There might be a number of objection to this patch. One
> I can think of is that this patch creates the nconnects based on
> whether or not the new mount asked for nconnect instead of
> unconditionally creating nconnect number of connections. The patch
> still creates nconnect connections based on the original value
> instead of picking the value of clp->cl_nconnect. I would have
> preferred that would be done. I don't see what can be wrong with
> using the new value. But I feared to go against what was objected
> before. My preference would be to (1) create clp->cl_nconnect
> connections or (2) not use this patch at all or (3) use as is
> here (meaning at least not create extra connections unless asked
> for by the mount).
> ---
>  fs/nfs/nfs4client.c | 9 +++++++++
>  1 file changed, 9 insertions(+)
>
> diff --git a/fs/nfs/nfs4client.c b/fs/nfs/nfs4client.c
> index af57332503be..50fa9d53b444 100644
> --- a/fs/nfs/nfs4client.c
> +++ b/fs/nfs/nfs4client.c
> @@ -427,6 +427,15 @@ static void nfs4_add_trunk(struct nfs_client *clp, struct nfs_client *old)
>
>         rpc_clnt_add_xprt(old->cl_rpcclient, &xprt_args,
>                           rpc_clnt_test_and_add_xprt, NULL);
> +
> +       if (clp->cl_nconnect > 1) {
> +               int i;
> +
> +               for (i = 0; i < old->cl_nconnect - 1; i++)
> +                       if (rpc_clnt_add_xprt(old->cl_rpcclient, &xprt_args,
> +                                             NULL, NULL) < 0)
> +                               break;
> +       }
>  }
>
>  /**
> --
> 2.27.0
>