----- Original Message -----
From: "Oleksandr Shchirskyi" <oleksandr.shchirskyi@xxxxxxxxxxxxxxx>
To: "Nigel Croxon" <ncroxon@xxxxxxxxxx>
Cc: linux-raid@xxxxxxxxxxxxxxx, "Mariusz Tkaczyk" <mariusz.tkaczyk@xxxxxxxxxxxxxxx>, "Jes Sorensen" <jes@xxxxxxxxxxxxxxxxxx>
Sent: Tuesday, March 23, 2021 4:58:27 PM
Subject: Re: [PATCH] mdadm: fix reshape from RAID5 to RAID6 with backup file
On 3/23/2021 5:36 PM, Nigel Croxon wrote:
> Oleksandr,
> Can you post your dmesg output when running the commands?
>
> I've back down from 5.11 to 5.8 and I still see:
> [ +0.042694] md/raid0:md126: raid5 must have missing parity disk!
> [ +0.000001] md: md126: raid0 would not accept array
>
> Thanks, Nigel
Hello Nigel,
I've switched to 4.18.0-240.el8.x86_64 kernel (I have RHEL8.3) and I still
have the same results, issue is still easily reproducible when patch
4ae96c8 is applied.
Cropped test logs with and w/o your patch:
# git log -n1 --oneline
f94df5c (HEAD -> master, origin/master, origin/HEAD) imsm: support for
third Sata controller
# make clean; make; make install-systemd; make install
# mdadm -CR imsm0 -e imsm -n4 /dev/nvme[0-3]n1 && mdadm -CR volume -l0
--chunk 64 --size=10G --raid-devices=1 /dev/nvme0n1 --force
# mdadm -G /dev/md/imsm0 -n2
# dmesg -c
[ 393.530389] md126: detected capacity change from 0 to 10737418240
[ 407.139318] md/raid:md126: device nvme0n1 operational as raid disk 0
[ 407.153920] md/raid:md126: raid level 4 active with 1 out of 2 devices,
algorithm 5
[ 407.246037] md: reshape of RAID array md126
[ 407.357940] md: md126: reshape interrupted.
[ 407.388144] md: reshape of RAID array md126
[ 407.398737] md: md126: reshape interrupted.
[ 407.403486] md: reshape of RAID array md126
[ 459.414250] md: md126: reshape done.
# cat /proc/mdstat
Personalities : [raid0] [raid6] [raid5] [raid4]
md126 : active raid4 nvme3n1[2] nvme0n1[0]
10485760 blocks super external:/md127/0 level 4, 64k chunk,
algorithm 0 [3/2] [UU_]
md127 : inactive nvme3n1[3](S) nvme2n1[2](S) nvme1n1[1](S) nvme0n1[0](S)
4420 blocks super external:imsm
unused devices: <none>
# mdadm -Ss; wipefs -a /dev/nvme[0-3]n1
# dmesg -C
# git revert 4ae96c802203ec3cfbb089240c56d61f7f4661b3
# make clean; make; make install-systemd; make install
# mdadm -CR imsm0 -e imsm -n4 /dev/nvme[0-3]n1 && mdadm -CR volume -l0
--chunk 64 --size=10G --raid-devices=1 /dev/nvme0n1 --force
# mdadm -G /dev/md/imsm0 -n2
# dmesg -c
[ 623.772039] md126: detected capacity change from 0 to 10737418240
[ 644.823245] md/raid:md126: device nvme0n1 operational as raid disk 0
[ 644.838542] md/raid:md126: raid level 4 active with 1 out of 2 devices,
algorithm 5
[ 644.928672] md: reshape of RAID array md126
[ 697.405351] md: md126: reshape done.
[ 697.409659] md126: detected capacity change from 10737418240 to 21474836480
# cat /proc/mdstat
Personalities : [raid0] [raid6] [raid5] [raid4]
md126 : active raid0 nvme3n1[2] nvme0n1[0]
20971520 blocks super external:/md127/0 64k chunks
md127 : inactive nvme3n1[3](S) nvme2n1[2](S) nvme1n1[1](S) nvme0n1[0](S)
4420 blocks super external:imsm
Do you need more detailed logs? My system/drives configuration details?
Regards,
Oleksandr Shchirskyi
>From f0c80c8e90b2ce113b6e22f919659430d3d20efa Mon Sep 17 00:00:00 2001
From: Nigel Croxon <ncroxon@xxxxxxxxxx>
Date: Fri, 26 Mar 2021 07:56:10 -0400
Subject: [PATCH] mdadm: fix growing containers
This fixes growing containers which was broken with
commit 4ae96c802203ec3c (mdadm: fix reshape from RAID5 to RAID6 with
backup file)
The issue being that containers use the function
wait_for_reshape_isms and expect a number value and not a
string value of "max". The change is to test for external
before setting the correct value.
Signed-off-by: Nigel Croxon <ncroxon@xxxxxxxxxx>
---
Grow.c | 19 +++++++++++--------
1 file changed, 11 insertions(+), 8 deletions(-)
diff --git a/Grow.c b/Grow.c
index 1120929..de28540 100644
--- a/Grow.c
+++ b/Grow.c
@@ -921,7 +921,7 @@ static int subarray_set_num(char *container, struct mdinfo *sra, char *name, int
}
int start_reshape(struct mdinfo *sra, int already_running,
- int before_data_disks, int data_disks)
+ int before_data_disks, int data_disks, struct supertype *st)
{
int err;
unsigned long long sync_max_to_set;
@@ -935,20 +935,23 @@ int start_reshape(struct mdinfo *sra, int already_running,
else
sync_max_to_set = (sra->component_size * data_disks
- sra->reshape_progress) / data_disks;
+
if (!already_running)
sysfs_set_num(sra, NULL, "sync_min", sync_max_to_set);
- err = err ?: sysfs_set_num(sra, NULL, "sync_max", sync_max_to_set);
+
+ if (st->ss->external)
+ err = err ?: sysfs_set_num(sra, NULL, "sync_max", sync_max_to_set);
+ else
+ err = err ?: sysfs_set_str(sra, NULL, "sync_max", "max");
+
if (!already_running && err == 0) {
int cnt = 5;
- int err2;
do {
err = sysfs_set_str(sra, NULL, "sync_action",
"reshape");
- err2 = sysfs_set_str(sra, NULL, "sync_max",
- "max");
- if (err || err2)
+ if (err)
sleep(1);
- } while (err && err2 && errno == EBUSY && cnt-- > 0);
+ } while (err && errno == EBUSY && cnt-- > 0);
}
return err;
}
@@ -3470,7 +3473,7 @@ started:
goto release;
err = start_reshape(sra, restart, reshape.before.data_disks,
- reshape.after.data_disks);
+ reshape.after.data_disks, st);
if (err) {
pr_err("Cannot %s reshape for %s\n",
restart ? "continue" : "start", devname);
--
2.27.0
