Subject: [PATCH 0/3] dm-mpath: dynamic load balancers (v1)

Subject: [PATCH 0/3] dm-mpath: dynamic load balancers (v1)

[Kiyoshi Ueda]

Hi,

The patch-set adds the following 2 dynamic load balancers:
  o dm-queue-length: queue-length oriented dynamic load balancer
  o dm-service-time: service-time oriented dynamic load balancer

This patch-set can be applied on top of 2.6.29-rc8.
No dependencies on Alasdair's linux-next patches.

NOTE:
While the patches compile and work with the current bio-based dm,
for them to *properly* work, the request-based dm patches should
be applied, too.
See <http://lkml.org/lkml/2008/9/12/100> for why request-based dm
improves multipath load balancing.

Summary of the patch-set:
  1/3: dm-mpath: interface change for dynamic load balancers
  2/3: dm-mpath: add queue-length oriented dynamic load balancer
  3/3: dm-mpath: add service-time oriented dynamic load balancer

 drivers/md/Kconfig            |   18 ++
 drivers/md/Makefile           |    2 
 drivers/md/dm-mpath.c         |   28 ++-
 drivers/md/dm-path-selector.h |    8 -
 drivers/md/dm-queue-length.c  |  257 +++++++++++++++++++++++++++++++++++
 drivers/md/dm-round-robin.c   |    2 
 drivers/md/dm-service-time.c  |  301 ++++++++++++++++++++++++++++++++++++++++++
 7 files changed, 603 insertions(+), 13 deletions(-)

Thanks,
Kiyoshi Ueda

[PATCH 1/3] dm-mpath: interface change for dynamic load balancers

[Kiyoshi Ueda]

This patch changes the dm path selector interface for dynamic
load balancers:
  o adding a new hook, start_io()
  o adding 'nr_bytes' parameter to select_path()/start_io()/end_io()
    to pass the size of the I/O

start_io() is called when a target driver actually submits I/O
to the selected path.
Path selectors can use it to start accounting of the I/O.
(e.g. counting the number of in-flight I/Os.)
The start_io hook is based on the patch posted by Stefan Bader:
https://www.redhat.com/archives/dm-devel/2005-October/msg00050.html

nr_bytes, the size of the I/O, is used by path selectors for
size-based decision.
dm-service-time uses it to estimate service time, for example.
(Added the nr_bytes member to dm_mpath_io instead of using existing
 details.bi_size, since request-based dm patch deletes it.)

Signed-off-by: Stefan Bader <stefan.bader@canonical.com>
Signed-off-by: Kiyoshi Ueda <k-ueda@ct.jp.nec.com>
Signed-off-by: Jun'ichi Nomura <j-nomura@ce.jp.nec.com>
Cc: Vijayakumar Balasubramanian <vijayakumar@hp.com>
Cc: Alasdair G Kergon <agk@redhat.com>
---
 drivers/md/dm-mpath.c         |   28 ++++++++++++++++++----------
 drivers/md/dm-path-selector.h |    8 ++++++--
 drivers/md/dm-round-robin.c   |    2 +-
 3 files changed, 25 insertions(+), 13 deletions(-)

Index: 2.6.29-rc8/drivers/md/dm-mpath.c
===================================================================
--- 2.6.29-rc8.orig/drivers/md/dm-mpath.c
+++ 2.6.29-rc8/drivers/md/dm-mpath.c
@@ -103,6 +103,7 @@ struct multipath {
 struct dm_mpath_io {
 	struct pgpath *pgpath;
 	struct dm_bio_details details;
+	size_t nr_bytes;
 };
 
 typedef int (*action_fn) (struct pgpath *pgpath);
@@ -251,11 +252,12 @@ static void __switch_pg(struct multipath
 	m->pg_init_count = 0;
 }
 
-static int __choose_path_in_pg(struct multipath *m, struct priority_group *pg)
+static int __choose_path_in_pg(struct multipath *m, struct priority_group *pg,
+			       size_t nr_bytes)
 {
 	struct dm_path *path;
 
-	path = pg->ps.type->select_path(&pg->ps, &m->repeat_count);
+	path = pg->ps.type->select_path(&pg->ps, &m->repeat_count, nr_bytes);
 	if (!path)
 		return -ENXIO;
 
@@ -267,7 +269,7 @@ static int __choose_path_in_pg(struct mu
 	return 0;
 }
 
-static void __choose_pgpath(struct multipath *m)
+static void __choose_pgpath(struct multipath *m, size_t nr_bytes)
 {
 	struct priority_group *pg;
 	unsigned bypassed = 1;
@@ -279,12 +281,12 @@ static void __choose_pgpath(struct multi
 	if (m->next_pg) {
 		pg = m->next_pg;
 		m->next_pg = NULL;
-		if (!__choose_path_in_pg(m, pg))
+		if (!__choose_path_in_pg(m, pg, nr_bytes))
 			return;
 	}
 
 	/* Don't change PG until it has no remaining paths */
-	if (m->current_pg && !__choose_path_in_pg(m, m->current_pg))
+	if (m->current_pg && !__choose_path_in_pg(m, m->current_pg, nr_bytes))
 		return;
 
 	/*
@@ -296,7 +298,7 @@ static void __choose_pgpath(struct multi
 		list_for_each_entry(pg, &m->priority_groups, list) {
 			if (pg->bypassed == bypassed)
 				continue;
-			if (!__choose_path_in_pg(m, pg))
+			if (!__choose_path_in_pg(m, pg, nr_bytes))
 				return;
 		}
 	} while (bypassed--);
@@ -327,6 +329,7 @@ static int map_io(struct multipath *m, s
 		  struct dm_mpath_io *mpio, unsigned was_queued)
 {
 	int r = DM_MAPIO_REMAPPED;
+	size_t nr_bytes = bio->bi_size;
 	unsigned long flags;
 	struct pgpath *pgpath;
 
@@ -335,7 +338,7 @@ static int map_io(struct multipath *m, s
 	/* Do we need to select a new pgpath? */
 	if (!m->current_pgpath ||
 	    (!m->queue_io && (m->repeat_count && --m->repeat_count == 0)))
-		__choose_pgpath(m);
+		__choose_pgpath(m, nr_bytes);
 
 	pgpath = m->current_pgpath;
 
@@ -360,6 +363,11 @@ static int map_io(struct multipath *m, s
 		r = -EIO;	/* Failed */
 
 	mpio->pgpath = pgpath;
+	mpio->nr_bytes = nr_bytes;
+
+	if (r == DM_MAPIO_REMAPPED && pgpath->pg->ps.type->start_io)
+		pgpath->pg->ps.type->start_io(&pgpath->pg->ps, &pgpath->path,
+					      nr_bytes);
 
 	spin_unlock_irqrestore(&m->lock, flags);
 
@@ -438,7 +446,7 @@ static void process_queued_ios(struct wo
 		goto out;
 
 	if (!m->current_pgpath)
-		__choose_pgpath(m);
+		__choose_pgpath(m, 0);
 
 	pgpath = m->current_pgpath;
 
@@ -1196,7 +1204,7 @@ static int multipath_end_io(struct dm_ta
 	if (pgpath) {
 		ps = &pgpath->pg->ps;
 		if (ps->type->end_io)
-			ps->type->end_io(ps, &pgpath->path);
+			ps->type->end_io(ps, &pgpath->path, mpio->nr_bytes);
 	}
 	if (r != DM_ENDIO_INCOMPLETE)
 		mempool_free(mpio, m->mpio_pool);
@@ -1412,7 +1420,7 @@ static int multipath_ioctl(struct dm_tar
 	spin_lock_irqsave(&m->lock, flags);
 
 	if (!m->current_pgpath)
-		__choose_pgpath(m);
+		__choose_pgpath(m, 0);
 
 	if (m->current_pgpath) {
 		bdev = m->current_pgpath->path.dev->bdev;
Index: 2.6.29-rc8/drivers/md/dm-path-selector.h
===================================================================
--- 2.6.29-rc8.orig/drivers/md/dm-path-selector.h
+++ 2.6.29-rc8/drivers/md/dm-path-selector.h
@@ -56,7 +56,8 @@ struct path_selector_type {
 	 * the path fails.
 	 */
 	struct dm_path *(*select_path) (struct path_selector *ps,
-				     unsigned *repeat_count);
+					unsigned *repeat_count,
+					size_t nr_bytes);
 
 	/*
 	 * Notify the selector that a path has failed.
@@ -75,7 +76,10 @@ struct path_selector_type {
 	int (*status) (struct path_selector *ps, struct dm_path *path,
 		       status_type_t type, char *result, unsigned int maxlen);
 
-	int (*end_io) (struct path_selector *ps, struct dm_path *path);
+	int (*start_io) (struct path_selector *ps, struct dm_path *path,
+			 size_t nr_bytes);
+	int (*end_io) (struct path_selector *ps, struct dm_path *path,
+		       size_t nr_bytes);
 };
 
 /* Register a path selector */
Index: 2.6.29-rc8/drivers/md/dm-round-robin.c
===================================================================
--- 2.6.29-rc8.orig/drivers/md/dm-round-robin.c
+++ 2.6.29-rc8/drivers/md/dm-round-robin.c
@@ -161,7 +161,7 @@ static int rr_reinstate_path(struct path
 }
 
 static struct dm_path *rr_select_path(struct path_selector *ps,
-				   unsigned *repeat_count)
+				      unsigned *repeat_count, size_t nr_bytes)
 {
 	struct selector *s = (struct selector *) ps->context;
 	struct path_info *pi = NULL;

[PATCH 2/3] dm-mpath: add queue-length oriented dynamic load balancer

[Kiyoshi Ueda]

This patch adds a dynamic load balancer, dm-queue-length, which
balances the number of in-flight I/Os.

The code is based on the patch posted by Stefan Bader:
https://www.redhat.com/archives/dm-devel/2005-October/msg00050.html


Signed-off-by: Stefan Bader <stefan.bader@canonical.com>
Signed-off-by: Kiyoshi Ueda <k-ueda@ct.jp.nec.com>
Signed-off-by: Jun'ichi Nomura <j-nomura@ce.jp.nec.com>
Cc: Vijayakumar Balasubramanian <vijayakumar@hp.com>
Cc: Alasdair G Kergon <agk@redhat.com>
---
 drivers/md/Kconfig           |    9 +
 drivers/md/Makefile          |    1 
 drivers/md/dm-queue-length.c |  257 +++++++++++++++++++++++++++++++++++++++++++
 3 files changed, 267 insertions(+)

Index: 2.6.29-rc8/drivers/md/dm-queue-length.c
===================================================================
--- /dev/null
+++ 2.6.29-rc8/drivers/md/dm-queue-length.c
@@ -0,0 +1,257 @@
+/*
+ * Copyright (C) 2004-2005 IBM Corp.  All Rights Reserved.
+ * Copyright (C) 2006-2009 NEC Corporation.
+ *
+ * dm-queue-length.c
+ *
+ * Module Author: Stefan Bader, IBM
+ * Modified by: Kiyoshi Ueda, NEC
+ *
+ * This file is released under the GPL.
+ *
+ * queue-length path selector, which chooses a path with the least number of
+ * in-flight I/Os.
+ */
+
+#include "dm.h"
+#include "dm-path-selector.h"
+
+#include <linux/slab.h>
+#include <linux/ctype.h>
+#include <linux/errno.h>
+#include <linux/module.h>
+#include <asm/atomic.h>
+
+#define DM_MSG_PREFIX	"multipath queue-length"
+#define QL_MIN_IO	128
+#define QL_VERSION	"0.1.0"
+
+struct selector {
+	struct list_head	valid_paths;
+	struct list_head	failed_paths;
+};
+
+struct path_info {
+	struct list_head	list;
+	struct dm_path		*path;
+	unsigned int		repeat_count;
+	atomic_t		qlen;
+};
+
+static struct selector *alloc_selector(void)
+{
+	struct selector *s = kmalloc(sizeof(*s), GFP_KERNEL);
+
+	if (s) {
+		INIT_LIST_HEAD(&s->valid_paths);
+		INIT_LIST_HEAD(&s->failed_paths);
+	}
+
+	return s;
+}
+
+static int ql_create(struct path_selector *ps, unsigned argc, char **argv)
+{
+	struct selector *s = alloc_selector();
+
+	if (!s)
+		return -ENOMEM;
+
+	ps->context = s;
+	return 0;
+}
+
+static void ql_free_paths(struct list_head *paths)
+{
+	struct path_info *pi, *next;
+
+	list_for_each_entry_safe(pi, next, paths, list) {
+		list_del(&pi->list);
+		kfree(pi);
+	}
+}
+
+static void ql_destroy(struct path_selector *ps)
+{
+	struct selector *s = (struct selector *) ps->context;
+
+	ql_free_paths(&s->valid_paths);
+	ql_free_paths(&s->failed_paths);
+	kfree(s);
+	ps->context = NULL;
+}
+
+static int ql_status(struct path_selector *ps, struct dm_path *path,
+		     status_type_t type, char *result, unsigned int maxlen)
+{
+	int sz = 0;
+	struct path_info *pi;
+
+	/* When called with (path == NULL), return selector status/args. */
+	if (!path)
+		DMEMIT("0 ");
+	else {
+		pi = path->pscontext;
+
+		switch (type) {
+		case STATUSTYPE_INFO:
+			DMEMIT("%u ", atomic_read(&pi->qlen));
+			break;
+		case STATUSTYPE_TABLE:
+			DMEMIT("%u ", pi->repeat_count);
+			break;
+		}
+	}
+
+	return sz;
+}
+
+static int ql_add_path(struct path_selector *ps, struct dm_path *path,
+		       int argc, char **argv, char **error)
+{
+	struct selector *s = (struct selector *) ps->context;
+	struct path_info *pi;
+	unsigned int repeat_count = QL_MIN_IO;
+
+	/* Parse the arguments */
+	if (argc > 1) {
+		*error = "queue-length ps: incorrect number of arguments";
+		return -EINVAL;
+	}
+
+	/* First path argument is number of I/Os before switching path. */
+	if ((argc == 1) && (sscanf(argv[0], "%u", &repeat_count) != 1)) {
+		*error = "queue-length ps: invalid repeat count";
+		return -EINVAL;
+	}
+
+	/* Allocate the path information structure */
+	pi = kmalloc(sizeof(*pi), GFP_KERNEL);
+	if (!pi) {
+		*error = "queue-length ps: Error allocating path information";
+		return -ENOMEM;
+	}
+
+	pi->path = path;
+	pi->repeat_count = repeat_count;
+	atomic_set(&pi->qlen, 0);
+
+	path->pscontext = pi;
+
+	list_add_tail(&pi->list, &s->valid_paths);
+
+	return 0;
+}
+
+static void ql_fail_path(struct path_selector *ps, struct dm_path *path)
+{
+	struct selector *s = (struct selector *) ps->context;
+	struct path_info *pi = path->pscontext;
+
+	list_move(&pi->list, &s->failed_paths);
+}
+
+static int ql_reinstate_path(struct path_selector *ps, struct dm_path *path)
+{
+	struct selector *s = (struct selector *) ps->context;
+	struct path_info *pi = path->pscontext;
+
+	list_move_tail(&pi->list, &s->valid_paths);
+
+	return 0;
+}
+
+static struct dm_path *ql_select_path(struct path_selector *ps,
+				      unsigned *repeat_count, size_t nr_bytes)
+{
+	struct selector *s = (struct selector *) ps->context;
+	struct path_info *pi = NULL, *best = NULL;
+
+	if (list_empty(&s->valid_paths))
+		return NULL;
+
+	/* Change preferred (first in list) path to evenly balance. */
+	list_move_tail(s->valid_paths.next, &s->valid_paths);
+
+	list_for_each_entry(pi, &s->valid_paths, list) {
+		if (!best ||
+		    (atomic_read(&pi->qlen) < atomic_read(&best->qlen)))
+			best = pi;
+
+		if (!atomic_read(&best->qlen))
+			break;
+	}
+
+	if (!best)
+		return NULL;
+
+	*repeat_count = best->repeat_count;
+
+	return best->path;
+}
+
+static int ql_start_io(struct path_selector *ps, struct dm_path *path,
+		       size_t nr_bytes)
+{
+	struct path_info *pi = path->pscontext;
+
+	atomic_inc(&pi->qlen);
+
+	return 0;
+}
+
+static int ql_end_io(struct path_selector *ps, struct dm_path *path,
+		     size_t nr_bytes)
+{
+	struct path_info *pi = path->pscontext;
+
+	atomic_dec(&pi->qlen);
+
+	return 0;
+}
+
+static struct path_selector_type ql_ps = {
+	.name		= "queue-length",
+	.module		= THIS_MODULE,
+	.table_args	= 1,
+	.info_args	= 1,
+	.create		= ql_create,
+	.destroy	= ql_destroy,
+	.status		= ql_status,
+	.add_path	= ql_add_path,
+	.fail_path	= ql_fail_path,
+	.reinstate_path	= ql_reinstate_path,
+	.select_path	= ql_select_path,
+	.start_io	= ql_start_io,
+	.end_io		= ql_end_io,
+};
+
+static int __init dm_ql_init(void)
+{
+	int r = dm_register_path_selector(&ql_ps);
+
+	if (r < 0)
+		DMERR("register failed %d", r);
+
+	DMINFO("version " QL_VERSION " loaded");
+
+	return r;
+}
+
+static void __exit dm_ql_exit(void)
+{
+	int r = dm_unregister_path_selector(&ql_ps);
+
+	if (r < 0)
+		DMERR("unregister failed %d", r);
+}
+
+module_init(dm_ql_init);
+module_exit(dm_ql_exit);
+
+MODULE_AUTHOR("Stefan Bader <Stefan.Bader at de.ibm.com>");
+MODULE_DESCRIPTION(
+	"(C) Copyright IBM Corp. 2004,2005   All Rights Reserved.\n"
+	DM_NAME " path selector to balance the number of in-flight I/Os"
+);
+MODULE_LICENSE("GPL");
Index: 2.6.29-rc8/drivers/md/Makefile
===================================================================
--- 2.6.29-rc8.orig/drivers/md/Makefile
+++ 2.6.29-rc8/drivers/md/Makefile
@@ -34,6 +34,7 @@ obj-$(CONFIG_BLK_DEV_DM)	+= dm-mod.o
 obj-$(CONFIG_DM_CRYPT)		+= dm-crypt.o
 obj-$(CONFIG_DM_DELAY)		+= dm-delay.o
 obj-$(CONFIG_DM_MULTIPATH)	+= dm-multipath.o dm-round-robin.o
+obj-$(CONFIG_DM_MULTIPATH_QL)	+= dm-queue-length.o
 obj-$(CONFIG_DM_SNAPSHOT)	+= dm-snapshot.o
 obj-$(CONFIG_DM_MIRROR)		+= dm-mirror.o dm-log.o dm-region-hash.o
 obj-$(CONFIG_DM_ZERO)		+= dm-zero.o
Index: 2.6.29-rc8/drivers/md/Kconfig
===================================================================
--- 2.6.29-rc8.orig/drivers/md/Kconfig
+++ 2.6.29-rc8/drivers/md/Kconfig
@@ -274,6 +274,15 @@ config DM_MULTIPATH
 	---help---
 	  Allow volume managers to support multipath hardware.
 
+config DM_MULTIPATH_QL
+	tristate "I/O Path Selector based on the number of in-flight I/Os"
+	depends on DM_MULTIPATH
+	---help---
+	  This path selector is a dynamic load balancer which selects
+	  a path having the least number of in-flight I/Os.
+
+	  If unsure, say N.
+
 config DM_DELAY
 	tristate "I/O delaying target (EXPERIMENTAL)"
 	depends on BLK_DEV_DM && EXPERIMENTAL

[PATCH 3/3] dm-mpath: add service-time oriented dynamic load balancer

[Kiyoshi Ueda]

This patch adds a service time oriented dynamic load balancer,
dm-service-time, which selects a path with the shortest estimated
service time for the incoming I/O.
The service time is estimated by dividing the in-flight I/O size
with performance value of each path.

The performance value can be given as a table argument at the table
loading time.  If no performance value is given, all paths are
recognized as equal performance.


Signed-off-by: Kiyoshi Ueda <k-ueda@ct.jp.nec.com>
Signed-off-by: Jun'ichi Nomura <j-nomura@ce.jp.nec.com>
---
 drivers/md/Kconfig           |    9 +
 drivers/md/Makefile          |    1 
 drivers/md/dm-service-time.c |  301 +++++++++++++++++++++++++++++++++++++++++++
 3 files changed, 311 insertions(+)

Index: 2.6.29-rc8/drivers/md/dm-service-time.c
===================================================================
--- /dev/null
+++ 2.6.29-rc8/drivers/md/dm-service-time.c
@@ -0,0 +1,301 @@
+/*
+ * Copyright (C) 2007-2009 NEC Corporation.  All Rights Reserved.
+ *
+ * Module Author: Kiyoshi Ueda
+ *
+ * This file is released under the GPL.
+ *
+ * Throughput oriented path selector.
+ */
+
+#include "dm.h"
+#include "dm-path-selector.h"
+
+#define DM_MSG_PREFIX	"multipath service-time"
+#define ST_MIN_IO	1
+#define ST_VERSION	"0.1.0"
+
+struct selector {
+	struct list_head valid_paths;
+	struct list_head failed_paths;
+};
+
+struct path_info {
+	struct list_head list;
+	struct dm_path *path;
+	unsigned int repeat_count;
+	size_t perf;
+	atomic_t in_flight_size;	/* Total size of in-flight I/Os */
+};
+
+static struct selector *alloc_selector(void)
+{
+	struct selector *s = kmalloc(sizeof(*s), GFP_KERNEL);
+
+	if (s) {
+		INIT_LIST_HEAD(&s->valid_paths);
+		INIT_LIST_HEAD(&s->failed_paths);
+	}
+
+	return s;
+}
+
+static int st_create(struct path_selector *ps, unsigned argc, char **argv)
+{
+	struct selector *s = alloc_selector();
+
+	if (!s)
+		return -ENOMEM;
+
+	ps->context = s;
+	return 0;
+}
+
+static void free_paths(struct list_head *paths)
+{
+	struct path_info *pi, *next;
+
+	list_for_each_entry_safe(pi, next, paths, list) {
+		list_del(&pi->list);
+		kfree(pi);
+	}
+}
+
+static void st_destroy(struct path_selector *ps)
+{
+	struct selector *s = (struct selector *) ps->context;
+
+	free_paths(&s->valid_paths);
+	free_paths(&s->failed_paths);
+	kfree(s);
+	ps->context = NULL;
+}
+
+static int st_status(struct path_selector *ps, struct dm_path *path,
+		     status_type_t type, char *result, unsigned int maxlen)
+{
+	int sz = 0;
+	struct path_info *pi;
+
+	if (!path)
+		DMEMIT("0 ");
+	else {
+		pi = path->pscontext;
+
+		switch (type) {
+		case STATUSTYPE_INFO:
+			DMEMIT("%u %lu ", atomic_read(&pi->in_flight_size),
+			       pi->perf);
+			break;
+		case STATUSTYPE_TABLE:
+			DMEMIT("%u %lu ", pi->repeat_count, pi->perf);
+			break;
+		}
+	}
+
+	return sz;
+}
+
+static int st_add_path(struct path_selector *ps, struct dm_path *path,
+		       int argc, char **argv, char **error)
+{
+	struct selector *s = (struct selector *) ps->context;
+	struct path_info *pi;
+	unsigned int repeat_count = ST_MIN_IO;
+	size_t perf = 1;
+
+	if (argc > 2) {
+		*error = "service-time ps: incorrect number of arguments";
+		return -EINVAL;
+	}
+
+	/* First path argument is number of I/Os before switching path. */
+	if ((argc > 0) && (sscanf(argv[0], "%u", &repeat_count) != 1)) {
+		*error = "service-time ps: invalid repeat count";
+		return -EINVAL;
+	}
+
+	/*
+	 * Second path argument is a relative performance value.
+	 * If 0 is given, the path isn't used while other paths having
+	 * a positive value are available.
+	 */
+	if ((argc == 2) && (sscanf(argv[1], "%lu", &perf) != 1)) {
+		*error = "service-time ps: invalid performance value";
+		return -EINVAL;
+	}
+
+	/* allocate the path */
+	pi = kmalloc(sizeof(*pi), GFP_KERNEL);
+	if (!pi) {
+		*error = "service-time ps: Error allocating path context";
+		return -ENOMEM;
+	}
+
+	pi->path = path;
+	pi->repeat_count = repeat_count;
+	pi->perf = perf;
+	atomic_set(&pi->in_flight_size, 0);
+
+	path->pscontext = pi;
+
+	list_add_tail(&pi->list, &s->valid_paths);
+
+	return 0;
+}
+
+static void st_fail_path(struct path_selector *ps, struct dm_path *path)
+{
+	struct selector *s = (struct selector *) ps->context;
+	struct path_info *pi = path->pscontext;
+
+	list_move(&pi->list, &s->failed_paths);
+}
+
+static int st_reinstate_path(struct path_selector *ps, struct dm_path *path)
+{
+	struct selector *s = (struct selector *) ps->context;
+	struct path_info *pi = path->pscontext;
+
+	list_move_tail(&pi->list, &s->valid_paths);
+
+	return 0;
+}
+
+/*
+ * Returns:
+ * < 0 : pi1 is better
+ * 0   : no difference between pi1 and pi2
+ * > 0 : pi2 is better
+ */
+static int st_compare_load(struct path_info *pi1, struct path_info *pi2,
+			   size_t incoming)
+{
+	size_t sz1, sz2;
+
+	sz1 = atomic_read(&pi1->in_flight_size);
+	sz2 = atomic_read(&pi2->in_flight_size);
+
+	/*
+	 * Case 1: Both have same performace value. Choose less loaded path.
+	 */
+	if (pi1->perf == pi2->perf)
+		return sz1 - sz2;
+
+	/*
+	 * Case 2a: Both have same load. Choose higher performance path.
+	 * Case 2b: One path has no performance value. Choose the other one.
+	 */
+	if (sz1 == sz2 || !pi1->perf || !pi2->perf)
+		return pi2->perf - pi1->perf;
+
+	/*
+	 * Case 3: Calculate service time. Choose faster path.
+	 *         if ((sz1+incoming)/pi1->perf < (sz2+incoming)/pi2->perf) pi1
+	 *         if ((sz1+incoming)/pi1->perf > (sz2+incoming)/pi2->perf) pi2
+	 */
+	sz1 += incoming;
+	sz2 += incoming;
+	while (sz1 && sz2 && (sz1 < pi1->perf) && (sz2 < pi2->perf)) {
+		/* Size is not big enough to compare by division. Shift up */
+		sz1 <<= 2;
+		sz2 <<= 2;
+	}
+	do_div(sz1, pi1->perf);
+	do_div(sz2, pi2->perf);
+
+	if (sz1 != sz2)
+		return sz1 - sz2;
+
+	/*
+	 * Case 4: Service time is equal. Choose higher performance path.
+	 */
+	return pi2->perf - pi1->perf;
+}
+
+static struct dm_path *st_select_path(struct path_selector *ps,
+				      unsigned *repeat_count, size_t nr_bytes)
+{
+	struct selector *s = (struct selector *) ps->context;
+	struct path_info *pi = NULL, *best = NULL;
+
+	if (list_empty(&s->valid_paths))
+		return NULL;
+
+	/* Change preferred (first in list) path to evenly balance. */
+	list_move_tail(s->valid_paths.next, &s->valid_paths);
+
+	list_for_each_entry(pi, &s->valid_paths, list)
+		if (!best || (st_compare_load(pi, best, nr_bytes) < 0))
+			best = pi;
+
+	if (!best)
+		return NULL;
+
+	*repeat_count = best->repeat_count;
+
+	return best->path;
+}
+
+static int st_start_io(struct path_selector *ps, struct dm_path *path,
+		       size_t nr_bytes)
+{
+	struct path_info *pi = path->pscontext;
+
+	atomic_add(nr_bytes, &pi->in_flight_size);
+
+	return 0;
+}
+
+static int st_end_io(struct path_selector *ps, struct dm_path *path,
+		     size_t nr_bytes)
+{
+	struct path_info *pi = path->pscontext;
+
+	atomic_sub(nr_bytes, &pi->in_flight_size);
+
+	return 0;
+}
+
+static struct path_selector_type st_ps = {
+	.name		= "service-time",
+	.module		= THIS_MODULE,
+	.table_args	= 2,
+	.info_args	= 2,
+	.create		= st_create,
+	.destroy	= st_destroy,
+	.status		= st_status,
+	.add_path	= st_add_path,
+	.fail_path	= st_fail_path,
+	.reinstate_path	= st_reinstate_path,
+	.select_path	= st_select_path,
+	.start_io	= st_start_io,
+	.end_io		= st_end_io,
+};
+
+static int __init dm_st_init(void)
+{
+	int r = dm_register_path_selector(&st_ps);
+
+	if (r < 0)
+		DMERR("register failed %d", r);
+
+	DMINFO("version " ST_VERSION " loaded");
+
+	return r;
+}
+
+static void __exit dm_st_exit(void)
+{
+	int r = dm_unregister_path_selector(&st_ps);
+
+	if (r < 0)
+		DMERR("unregister failed %d", r);
+}
+
+module_init(dm_st_init);
+module_exit(dm_st_exit);
+
+MODULE_DESCRIPTION(DM_NAME " throughput oriented path selector");
+MODULE_AUTHOR("Kiyoshi Ueda <k-ueda@ct.jp.nec.com>");
+MODULE_LICENSE("GPL");
Index: 2.6.29-rc8/drivers/md/Makefile
===================================================================
--- 2.6.29-rc8.orig/drivers/md/Makefile
+++ 2.6.29-rc8/drivers/md/Makefile
@@ -35,6 +35,7 @@ obj-$(CONFIG_DM_CRYPT)		+= dm-crypt.o
 obj-$(CONFIG_DM_DELAY)		+= dm-delay.o
 obj-$(CONFIG_DM_MULTIPATH)	+= dm-multipath.o dm-round-robin.o
 obj-$(CONFIG_DM_MULTIPATH_QL)	+= dm-queue-length.o
+obj-$(CONFIG_DM_MULTIPATH_ST)	+= dm-service-time.o
 obj-$(CONFIG_DM_SNAPSHOT)	+= dm-snapshot.o
 obj-$(CONFIG_DM_MIRROR)		+= dm-mirror.o dm-log.o dm-region-hash.o
 obj-$(CONFIG_DM_ZERO)		+= dm-zero.o
Index: 2.6.29-rc8/drivers/md/Kconfig
===================================================================
--- 2.6.29-rc8.orig/drivers/md/Kconfig
+++ 2.6.29-rc8/drivers/md/Kconfig
@@ -283,6 +283,15 @@ config DM_MULTIPATH_QL
 
 	  If unsure, say N.
 
+config DM_MULTIPATH_ST
+	tristate "I/O Path Selector based on the service time"
+	depends on DM_MULTIPATH
+	---help---
+	  This path selector is a dynamic load balancer which selects
+	  a path to complete the incoming I/O with the shortest time.
+
+	  If unsure, say N.
+
 config DM_DELAY
 	tristate "I/O delaying target (EXPERIMENTAL)"
 	depends on BLK_DEV_DM && EXPERIMENTAL

[PATCH 1/3] dm-mpath: interface change for dynamic load balancers

[Kiyoshi Ueda]

This patch changes the dm path selector interface for dynamic
load balancers:
  o adding a new hook, start_io()
  o adding 'nr_bytes' parameter to select_path()/start_io()/end_io()
    to pass the size of the I/O

start_io() is called when a target driver actually submits I/O
to the selected path.
Path selectors can use it to start accounting of the I/O.
(e.g. counting the number of in-flight I/Os.)
The start_io hook is based on the patch posted by Stefan Bader:
https://www.redhat.com/archives/dm-devel/2005-October/msg00050.html

nr_bytes, the size of the I/O, is used by path selectors for
size-based decision.
dm-service-time uses it to estimate service time, for example.
(Added the nr_bytes member to dm_mpath_io instead of using existing
 details.bi_size, since request-based dm patch deletes it.)

Signed-off-by: Stefan Bader <stefan.bader@canonical.com>
Signed-off-by: Kiyoshi Ueda <k-ueda@ct.jp.nec.com>
Signed-off-by: Jun'ichi Nomura <j-nomura@ce.jp.nec.com>
Cc: Alasdair G Kergon <agk@redhat.com>
---
 drivers/md/dm-mpath.c         |   28 ++++++++++++++++++----------
 drivers/md/dm-path-selector.h |    8 ++++++--
 drivers/md/dm-round-robin.c   |    2 +-
 3 files changed, 25 insertions(+), 13 deletions(-)

Index: 2.6.30-rc3/drivers/md/dm-mpath.c
===================================================================
--- 2.6.30-rc3.orig/drivers/md/dm-mpath.c
+++ 2.6.30-rc3/drivers/md/dm-mpath.c
@@ -102,6 +102,7 @@ struct multipath {
 struct dm_mpath_io {
 	struct pgpath *pgpath;
 	struct dm_bio_details details;
+	size_t nr_bytes;
 };
 
 typedef int (*action_fn) (struct pgpath *pgpath);
@@ -250,11 +251,12 @@ static void __switch_pg(struct multipath
 	m->pg_init_count = 0;
 }
 
-static int __choose_path_in_pg(struct multipath *m, struct priority_group *pg)
+static int __choose_path_in_pg(struct multipath *m, struct priority_group *pg,
+			       size_t nr_bytes)
 {
 	struct dm_path *path;
 
-	path = pg->ps.type->select_path(&pg->ps, &m->repeat_count);
+	path = pg->ps.type->select_path(&pg->ps, &m->repeat_count, nr_bytes);
 	if (!path)
 		return -ENXIO;
 
@@ -266,7 +268,7 @@ static int __choose_path_in_pg(struct mu
 	return 0;
 }
 
-static void __choose_pgpath(struct multipath *m)
+static void __choose_pgpath(struct multipath *m, size_t nr_bytes)
 {
 	struct priority_group *pg;
 	unsigned bypassed = 1;
@@ -278,12 +280,12 @@ static void __choose_pgpath(struct multi
 	if (m->next_pg) {
 		pg = m->next_pg;
 		m->next_pg = NULL;
-		if (!__choose_path_in_pg(m, pg))
+		if (!__choose_path_in_pg(m, pg, nr_bytes))
 			return;
 	}
 
 	/* Don't change PG until it has no remaining paths */
-	if (m->current_pg && !__choose_path_in_pg(m, m->current_pg))
+	if (m->current_pg && !__choose_path_in_pg(m, m->current_pg, nr_bytes))
 		return;
 
 	/*
@@ -295,7 +297,7 @@ static void __choose_pgpath(struct multi
 		list_for_each_entry(pg, &m->priority_groups, list) {
 			if (pg->bypassed == bypassed)
 				continue;
-			if (!__choose_path_in_pg(m, pg))
+			if (!__choose_path_in_pg(m, pg, nr_bytes))
 				return;
 		}
 	} while (bypassed--);
@@ -326,6 +328,7 @@ static int map_io(struct multipath *m, s
 		  struct dm_mpath_io *mpio, unsigned was_queued)
 {
 	int r = DM_MAPIO_REMAPPED;
+	size_t nr_bytes = bio->bi_size;
 	unsigned long flags;
 	struct pgpath *pgpath;
 
@@ -334,7 +337,7 @@ static int map_io(struct multipath *m, s
 	/* Do we need to select a new pgpath? */
 	if (!m->current_pgpath ||
 	    (!m->queue_io && (m->repeat_count && --m->repeat_count == 0)))
-		__choose_pgpath(m);
+		__choose_pgpath(m, nr_bytes);
 
 	pgpath = m->current_pgpath;
 
@@ -359,6 +362,11 @@ static int map_io(struct multipath *m, s
 		r = -EIO;	/* Failed */
 
 	mpio->pgpath = pgpath;
+	mpio->nr_bytes = nr_bytes;
+
+	if (r == DM_MAPIO_REMAPPED && pgpath->pg->ps.type->start_io)
+		pgpath->pg->ps.type->start_io(&pgpath->pg->ps, &pgpath->path,
+					      nr_bytes);
 
 	spin_unlock_irqrestore(&m->lock, flags);
 
@@ -437,7 +445,7 @@ static void process_queued_ios(struct wo
 		goto out;
 
 	if (!m->current_pgpath)
-		__choose_pgpath(m);
+		__choose_pgpath(m, 0);
 
 	pgpath = m->current_pgpath;
 
@@ -1195,7 +1203,7 @@ static int multipath_end_io(struct dm_ta
 	if (pgpath) {
 		ps = &pgpath->pg->ps;
 		if (ps->type->end_io)
-			ps->type->end_io(ps, &pgpath->path);
+			ps->type->end_io(ps, &pgpath->path, mpio->nr_bytes);
 	}
 	if (r != DM_ENDIO_INCOMPLETE)
 		mempool_free(mpio, m->mpio_pool);
@@ -1411,7 +1419,7 @@ static int multipath_ioctl(struct dm_tar
 	spin_lock_irqsave(&m->lock, flags);
 
 	if (!m->current_pgpath)
-		__choose_pgpath(m);
+		__choose_pgpath(m, 0);
 
 	if (m->current_pgpath) {
 		bdev = m->current_pgpath->path.dev->bdev;
Index: 2.6.30-rc3/drivers/md/dm-path-selector.h
===================================================================
--- 2.6.30-rc3.orig/drivers/md/dm-path-selector.h
+++ 2.6.30-rc3/drivers/md/dm-path-selector.h
@@ -56,7 +56,8 @@ struct path_selector_type {
 	 * the path fails.
 	 */
 	struct dm_path *(*select_path) (struct path_selector *ps,
-				     unsigned *repeat_count);
+					unsigned *repeat_count,
+					size_t nr_bytes);
 
 	/*
 	 * Notify the selector that a path has failed.
@@ -75,7 +76,10 @@ struct path_selector_type {
 	int (*status) (struct path_selector *ps, struct dm_path *path,
 		       status_type_t type, char *result, unsigned int maxlen);
 
-	int (*end_io) (struct path_selector *ps, struct dm_path *path);
+	int (*start_io) (struct path_selector *ps, struct dm_path *path,
+			 size_t nr_bytes);
+	int (*end_io) (struct path_selector *ps, struct dm_path *path,
+		       size_t nr_bytes);
 };
 
 /* Register a path selector */
Index: 2.6.30-rc3/drivers/md/dm-round-robin.c
===================================================================
--- 2.6.30-rc3.orig/drivers/md/dm-round-robin.c
+++ 2.6.30-rc3/drivers/md/dm-round-robin.c
@@ -161,7 +161,7 @@ static int rr_reinstate_path(struct path
 }
 
 static struct dm_path *rr_select_path(struct path_selector *ps,
-				   unsigned *repeat_count)
+				      unsigned *repeat_count, size_t nr_bytes)
 {
 	struct selector *s = (struct selector *) ps->context;
 	struct path_info *pi = NULL;