[Solved] Red lights blinking after setting up RAID0

[esousra]

I have 2 x 3TB disks on my 2Big Network 2

After setting them up as RAID0, every 3-4 secs the back disk lights, and the front eye, blink red.

There was no error during the disk setup process; and on the GUI I don't see anything wrong ("No Alerts", and SMART status are Ok).

After reboot it continues the same.

Any idea what can be wrong? Any log files I can check? - appreciate any help.

Attaching a couple of screenshots from my system.

[Jocko]

Hi

I suppose you are aware that with a raid0 you will lose all your data on a disk failure event. There is no software to recover these on a broken raid0. You should use a backup solution for your valuable data.

Otherwise all seems to be ok except for LED status.

Chek if a mdadm process for monitoring your raid is running

Code: Select all

ps ax|grep mdadm

This one checks your raid status every 10 min and updates the LED status or sends a mail (if you have enabled email notification feature)
Sometime after rebooting it fails to set rightly the LED status and you need to wait this delay so that you get the right LED status...

Post also these outputs

Code: Select all

cat /etc/mdadm.conf

and check if you run this command, red blinking stops

Code: Select all

/usr/bin/mdadm-events

if not that would mean the command detects wrongly the raid status with the raid0 level and then post

Code: Select all

sh -x /usr/bin/mdadm-events

[esousra]

Hi, thanks for quick feedback :)

I suppose you are aware that with a raid0 you will lose all your data on a disk failure event. There is no software to recover these on a broken raid0. You should use a backup solution for your valuable data.

Yap. I have a 2Big NAS as well, also with 2 x 3TB, in RAID1 (Seagate NAS OS 4), which I've been using for many years and is almost full. My purpose with this 2nd LaCie is to have them both in RAID0 and rsync'ed, on different "sites"; that way earning in capacity and geo redundancy :)

About the commands you suggested:

Code: Select all

root@fvdwsl-base:/ # ps ax|grep mdadm
 1368 ?        Ss     0:00 /sbin/mdadm --monitor --daemonise --scan --delay=600 --pid-file=/var/run/mdadm.pid
 2486 pts/0    S+     0:00 grep mdadm

Code: Select all

root@fvdwsl-base:/ # cat /etc/mdadm.conf
CREATE owner=root group=root mode=0666 auto=yes metadata=1.0
PROGRAM /usr/bin/mdadm-events
DEVICE /dev/sd* /dev/se*

#MAILADDR

ARRAY /dev/md0 metadata=1.0 level=raid0 num-devices=2 UUID=b0f15618:2b440fec:6c05b24e:09e36e57

and check if you run this command, red blinking stops
Code:
/usr/bin/mdadm-events

I tried it, and no, it doesn't stop. This was the output:

Code: Select all

root@fvdwsl-base:/ # /usr/bin/mdadm-events
USER=root
HOSTNAME=fvdwsl-base
HOME=/
SSH_TTY=/dev/pts/0
LOGNAME=root
RSYNC_RSH=/usr/bin/dbclient -p 22
TERM=xterm
PATH=/usr/bin:/bin:/sbin:/usr/sbin:/usr/local/bin:/usr/local/sbin:/opt/bin:/opt/sbin:/opt/usr/bin:/opt/usr/sbin:/opt/lib
SHELL=/bin/sh
PWD=/
LC_ALL=en_US.utf8
SSH_CONNECTION=192.168.10.2 55683 192.168.0.4 22
isVM=0
cat: /sys/block/md0/md/sync_action: No such file or directory
sh: resync: unknown operand


Last command output:

Code: Select all

root@fvdwsl-base:/ # sh -x /usr/bin/mdadm-events
+ grep -ic hypervisor /proc/cpuinfo
+ export isVM=0
+ [ -z  ]
+ . /etc/leds_name_table
+ head -1 /etc/platform
+ NAS=2big2
+ [ -h /sys/class/leds/ns_lite_v2:blue:sata ]
+ BFILE=/rw_fs/etc/led-brightness
+ MaxLevel=255
+ MaxLevel=7
+ frontred=net2big-v2:red:power
+ frontblue=net2big-v2:blue:power
+ sata0blue=net2big-v2:blue:sata0
+ sata0red=net2big-v2:red:sata0
+ sata1blue=net2big-v2:blue:sata1
+ sata1red=net2big-v2:red:sata1
+ [ -f /rw_fs/etc/led-brightness ]
+ brightness_level=7
+ env
USER=root
HOSTNAME=fvdwsl-base
HOME=/
SSH_TTY=/dev/pts/0
LOGNAME=root
RSYNC_RSH=/usr/bin/dbclient -p 22
TERM=xterm
PATH=/usr/bin:/bin:/sbin:/usr/sbin:/usr/local/bin:/usr/local/sbin:/opt/bin:/opt/sbin:/opt/usr/bin:/opt/usr/sbin:/opt/lib
SHELL=/bin/sh
PWD=/
LC_ALL=en_US.utf8
SSH_CONNECTION=192.168.10.2 55683 192.168.0.4 22
isVM=0
+ [ -f /tmp/booting ]
+ [  == force ]
+ force=0
+ /usr/bin/led
+ [ 7 -ne 0 ]
+ isoff=0
+ ALERTFILE=/rw_fs/tmp/raidfailure
+ hasdegraded=0
+ hasfaulty=0
+ /usr/bin/awk / md[0-9]/ {print $4}
+ /usr/bin/cat /proc/partitions
+ cat /sys/block/md0/md/array_state
+ status=clean
+ [ -f /sys/block/md0/md/degraded ]
+ degraded=0
+ devices=
+ ls -d /sys/block/md0/md/dev-sda8 /sys/block/md0/md/dev-sdb8
+ devices=sda8
+ devices=sda8 sdb8
+ [ clean == clean ]
+ [ 0 -eq 0 ]
+ cat /sys/block/md0/md/sync_action
cat: /sys/block/md0/md/sync_action: No such file or directory
+ [ != resync ]
sh: resync: unknown operand
+ hasdegraded=1
+ cat /sys/block/md0/md/dev-sda8/state
+ devstate=in_sync
+ [ in_sync == in_sync ]
+ [ 0 = 0 ]
+ printf %d 'a
+ expr 97 - 97
+ led=sata0
+ [ 0 -eq 0 ]
+ eval echo $sata0blue
+ echo net2big-v2:blue:sata0
+ [ -n net2big-v2:blue:sata0 ]
+ delayoff=2
+ eval echo $sata0red
+ echo net2big-v2:red:sata0
+ [ -n net2big-v2:red:sata0 ]
+ /usr/bin/led -l sata0red -w on
+ /usr/bin/led -l sata0red -o trigger -w timer
+ /usr/bin/led -l sata0red -o delay_on -w 4000
+ /usr/bin/led -l sata0red -o delay_off -w 2
+ [ 0 -eq 0 ]
+ eval echo $sata0blue
+ echo net2big-v2:blue:sata0
+ [ -n net2big-v2:blue:sata0 ]
+ /usr/bin/led -l sata0blue -w on
+ /usr/bin/led -l sata0blue -o trigger -w timer
+ /usr/bin/led -l sata0blue -o delay_on -w 3000
+ /usr/bin/led -l sata0blue -o delay_off -w 2
+ eval echo $sata0blue
+ echo net2big-v2:blue:sata0
+ [ -f /sys/class/leds/net2big-v2:blue:sata0/sata ]
+ /usr/bin/led -l sata0blue -o sata -w 0
+ cat /sys/block/md0/md/dev-sdb8/state
+ devstate=in_sync
+ [ in_sync == in_sync ]
+ [ 0 = 0 ]
+ printf %d 'b
+ expr 98 - 97
+ led=sata1
+ [ 0 -eq 0 ]
+ eval echo $sata1blue
+ echo net2big-v2:blue:sata1
+ [ -n net2big-v2:blue:sata1 ]
+ delayoff=2
+ eval echo $sata1red
+ echo net2big-v2:red:sata1
+ [ -n net2big-v2:red:sata1 ]
+ /usr/bin/led -l sata1red -w on
+ /usr/bin/led -l sata1red -o trigger -w timer
+ /usr/bin/led -l sata1red -o delay_on -w 4000
+ /usr/bin/led -l sata1red -o delay_off -w 2
+ [ 0 -eq 0 ]
+ eval echo $sata1blue
+ echo net2big-v2:blue:sata1
+ [ -n net2big-v2:blue:sata1 ]
+ /usr/bin/led -l sata1blue -w on
+ /usr/bin/led -l sata1blue -o trigger -w timer
+ /usr/bin/led -l sata1blue -o delay_on -w 3000
+ /usr/bin/led -l sata1blue -o delay_off -w 2
+ eval echo $sata1blue
+ echo net2big-v2:blue:sata1
+ [ -f /sys/class/leds/net2big-v2:blue:sata1/sata ]
+ /usr/bin/led -l sata1blue -o sata -w 0
+ [ 1 -eq 1 ]
+ [ 0 = 0 ]
+ [ -n net2big-v2:red:power ]
+ /usr/bin/led -l frontred -w on
+ /usr/bin/led -l frontred -o trigger -w timer
+ /usr/bin/led -l frontred -o delay_on -w 3000
+ [ 0 -eq 0 ]
+ /usr/bin/led -l frontred -o delay_off -w 5
+ [ 0 -eq 0 ]
+ /usr/bin/led -w on
+ /usr/bin/led -o trigger -w timer
+ /usr/bin/led -o delay_on -w 3000
+ /usr/bin/led -o delay_off -w 5
+ [ 0 -eq 1 ]
+ unlink /rw_fs/tmp/raidfailure
root@fvdwsl-base:/ #


Hope it helps?

[Jocko]

Thank you

Yes it helped me and it is an issue with the command "mdadm-events"

Please to post this output, I need it for patching the command

Code: Select all

cd /sys/block/md0/md
grep '' ./*

[esousra]

:) Here you go:

Code: Select all

root@fvdwsl-base:/ # cd /sys/block/md0/md
root@fvdwsl-base:/sys/block/md0/md # grep '' ./*
./array_size:default
./array_state:clean
./chunk_size:524288
./component_size:0
./consistency_policy:none
./layout:0
./level:raid0
./max_read_errors:20
./metadata_version:1.0
grep: ./new_dev: Permission denied
./raid_disks:2
./reshape_direction:forwards
./reshape_position:none
./resync_start:none
./safe_mode_delay:0.210
root@fvdwsl-base:/sys/block/md0/md #

[Jocko]

So

For testing before including it in a next patch, a new version of the command.

So :
- download the tarball (use the link sent by pm)
- store it in a share (I assume fvdw)
- from a shell window, run the tar command

Code: Select all

tar -xf /direct-usb/fvdw/fvdw-sl-18-2-raid-led-fix-29jan2020.tgz -C /

Then check if the led issue is fixed

Code: Select all

/usr/bin/mdadm-events

Now red blinking should be off :whistle

[esousra]

Aaand indeed it is, fixed :-D Thank you very much Jocko!

[Jocko]

So we will add the new version in the next patch

[MacGTO]

Hi,
i have the same issue.
can i have the same tar?

[Jocko]

Hi MacGTO

I just sent you a pm