Skip to contentSkip to conversation

force parallel resync

PATCH New Bernd Schubert 1 participant 1 message Dec 3, 2007 MD RAID

lore
Share & tools
lore

Bernd Schubert

author
#0 lore raw

PATCH 1/1 force parallel resync

Hello Neil,

on doing software-raid over some Infortrend hardware raid boxes, we can easily bring a single cpu to 100%, while the hardware-raid boxes are still only utilized for 50% or less.
So we are using *several* software raids instead of only one. However, for us it then also doesn't make sense to have a resync for only one of the md-raids. Below is a patch to force parallel resync for specified md-sets via sysfs.
This still leaves a problem, though. If a resync starts before settings this parameter, I don't know how to wake up the sleeping resync-thread. Any ideas?

Signed-off-by Bernd Schubert <[email protected]>

(The patch is against 2.6.22, but also applies to 2.6.23).

drivers/md/md.c +28 -2
Index: linux-2.6.22/drivers/md/md.c
===================================================================
@@ -2843,6 +2843,30 @@ __ATTR(sync_speed_max, S_IRUGO|S_IWUSR, 
 
 
 static ssize_t
+sync_force_parallel_show(mddev_t *mddev, char *page)
+{
+        return sprintf(page, "%d\n", mddev->parallel_resync);
+}
+
+static ssize_t
+sync_force_parallel_store(mddev_t *mddev, const char *buf, size_t len)
+{
+	char *e;
+	unsigned long n = simple_strtoul(buf, &e, 10);
+
+	if (!*buf || (*e && *e != '\n') || (n != 0 && n != 1))
+		return -EINVAL;
+
+	mddev->parallel_resync = n;
+	return len;
+}
+
+/* force parallel resync, even with shared block devices */
+static struct md_sysfs_entry md_sync_force_parallel =
+__ATTR(sync_force_parallel, S_IRUGO|S_IWUSR,
+       sync_force_parallel_show, sync_force_parallel_store);
+
+static ssize_t
 sync_speed_show(mddev_t *mddev, char *page)
 {
 	unsigned long resync, dt, db;

@@ -2980,6 +3004,7 @@ static struct attribute *md_redundancy_a
 	&md_sync_min.attr,
 	&md_sync_max.attr,
 	&md_sync_speed.attr,
+	&md_sync_force_parallel.attr,
 	&md_sync_completed.attr,
 	&md_suspend_lo.attr,
 	&md_suspend_hi.attr,

@@ -5264,8 +5289,9 @@ void md_do_sync(mddev_t *mddev)
 		ITERATE_MDDEV(mddev2,tmp) {
 			if (mddev2 == mddev)
 				continue;
-			if (mddev2->curr_resync && 
-			    match_mddev_units(mddev,mddev2)) {
+			if (!mddev->parallel_resync
+			&&  mddev2->curr_resync
+			&&  match_mddev_units(mddev,mddev2)) {
 				DEFINE_WAIT(wq);
 				if (mddev < mddev2 && mddev->curr_resync == 2) {
 					/* arbitrarily yield */

Index: linux-2.6.22/include/linux/raid/md_k.h
===================================================================
--- linux-2.6.22.orig/include/linux/raid/md_k.h 2007-12-03 13:00:10.000000000
+0100
+++ linux-2.6.22/include/linux/raid/md_k.h 2007-12-03 13:01:40.000000000 +0100
@@ -170,6 +170,9 @@ struct mddev_s
int sync_speed_min;
int sync_speed_max;

+ /* resync even though the same disks are shared among md-devices */
+ int parallel_resync;
+
int ok_start_degraded;
/* recovery/resync flags
* NEEDED: we might need to start a resync/recover

Thanks,
Bernd

Signature · 2 lines
Bernd Schubert
Q-Leap Networks GmbH