From: BingJing Chang <bingjingc@xxxxxxxxxxxx>
When any sync action is finishing or interrupted, the progress will
show 99.9% before the sync thread is reaped. Many reporters has asked
what happened to the last blocks. It might be a confusing meaning for
users because the progress will be backward after the interrupted task
is restarted.
Take a raid5 reshape for example:
mdadm -C --assume-clean /dev/md0 -l5 -n3 /dev/loop[012]
echo 2000 > /proc/sys/dev/raid/speed_limit_max
echo 1000 > /proc/sys/dev/raid/speed_limit_min # slow down the speed
mdadm /dev/md0 -a /dev/loop3
mdadm /dev/md0 --grow -n4
while true
do
mdadm -S /dev/md0
sleep 3
mdadm -A /dev/md0 /dev/loop[0123]
done
And you can see the fake 99.9% progress by the following command:
> while true; do cat /proc/mdstat | grep reshape; done
This confusing state can be fixed by exposing the real state to users.
And I also correct the sync action type for display.
Reported-by: Edwin Lin <edwinlin@xxxxxxxxxxxx>
Reviewed-by: Allen Peng <allenpeng@xxxxxxxxxxxx>
Signed-off-by: BingJing Chang <bingjingc@xxxxxxxxxxxx>
---
drivers/md/md.c | 25 +++++++++++++++----------
1 file changed, 15 insertions(+), 10 deletions(-)
diff --git a/drivers/md/md.c b/drivers/md/md.c
index bc67ab6..6714ced 100644
--- a/drivers/md/md.c
+++ b/drivers/md/md.c
@@ -7607,6 +7607,14 @@ static int status_resync(struct seq_file *seq, struct mddev *mddev)
sector_t rt;
int scale;
unsigned int per_milli;
+ char *sync_action;
+
+ sync_action = (test_bit(MD_RECOVERY_RESHAPE, &mddev->recovery) ?
+ "reshape" :
+ (test_bit(MD_RECOVERY_CHECK, &mddev->recovery) ?
+ "check" :
+ (test_bit(MD_RECOVERY_SYNC, &mddev->recovery) ?
+ "resync" : "recovery")));
if (test_bit(MD_RECOVERY_SYNC, &mddev->recovery) ||
test_bit(MD_RECOVERY_RESHAPE, &mddev->recovery))
@@ -7616,9 +7624,11 @@ static int status_resync(struct seq_file *seq, struct mddev *mddev)
resync = mddev->curr_resync;
if (resync <= 3) {
- if (test_bit(MD_RECOVERY_DONE, &mddev->recovery))
+ if (test_bit(MD_RECOVERY_DONE, &mddev->recovery)) {
/* Still cleaning up */
- resync = max_sectors;
+ seq_printf(seq, "\t%s=CLEANING UP", sync_action);
+ return 1;
+ }
} else if (resync > max_sectors)
resync = max_sectors;
else
@@ -7626,13 +7636,13 @@ static int status_resync(struct seq_file *seq, struct mddev *mddev)
if (resync == 0) {
if (mddev->recovery_cp < MaxSector) {
- seq_printf(seq, "\tresync=PENDING");
+ seq_printf(seq, "\t%s=PENDING", sync_action);
return 1;
}
return 0;
}
if (resync < 3) {
- seq_printf(seq, "\tresync=DELAYED");
+ seq_printf(seq, "\t%s=DELAYED", sync_action);
return 1;
}
@@ -7662,12 +7672,7 @@ static int status_resync(struct seq_file *seq, struct mddev *mddev)
seq_printf(seq, "] ");
}
seq_printf(seq, " %s =%3u.%u%% (%llu/%llu)",
- (test_bit(MD_RECOVERY_RESHAPE, &mddev->recovery)?
- "reshape" :
- (test_bit(MD_RECOVERY_CHECK, &mddev->recovery)?
- "check" :
- (test_bit(MD_RECOVERY_SYNC, &mddev->recovery) ?
- "resync" : "recovery"))),
+ sync_action,
per_milli/10, per_milli % 10,
(unsigned long long) resync/2,
(unsigned long long) max_sectors/2);
--
2.7.4
--
To unsubscribe from this list: send the line "unsubscribe linux-raid" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
