spin down discs

[matzi]

I have installed two discs, one in slot 1 one in slot 2. One or two of the discs is not spinning down even after hours. I can see the blue LED of slot 1 showing some action. I have not activated any media servers nor cron daemon.
Attached is my ps-A. Do you have any idea what is preventing the spin down?

root@fvdwsl-base:/ # ps -A
PID TTY TIME CMD
1 ? 00:00:01 init
2 ? 00:00:00 kthreadd
3 ? 00:00:00 ksoftirqd/0
4 ? 00:00:00 kworker/0:0
5 ? 00:00:00 kworker/0:0H
6 ? 00:00:00 kworker/u2:0
7 ? 00:00:00 rcu_preempt
8 ? 00:00:00 rcu_sched
9 ? 00:00:00 rcu_bh
10 ? 00:00:00 khelper
11 ? 00:00:00 kworker/u2:1
235 ? 00:00:00 writeback
237 ? 00:00:00 bioset
239 ? 00:00:00 kblockd
244 ? 00:00:00 ata_sff
254 ? 00:00:00 khubd
261 ? 00:00:00 md
271 ? 00:00:00 cfg80211
272 ? 00:00:00 kworker/0:1
287 ? 00:00:00 rpciod
296 ? 00:00:00 kswapd0
297 ? 00:00:00 fsnotify_mark
298 ? 00:00:00 nfsiod
299 ? 00:00:00 cifsiod
303 ? 00:00:00 jfsIO
304 ? 00:00:00 jfsCommit
305 ? 00:00:00 jfsSync
306 ? 00:00:00 xfsalloc
307 ? 00:00:00 xfs_mru_cache
308 ? 00:00:00 xfslogd
309 ? 00:00:00 crypto
432 ? 00:00:00 scsi_eh_0
435 ? 00:00:00 scsi_eh_1
438 ? 00:00:00 kworker/u2:2
439 ? 00:00:00 kworker/u2:3
445 ? 00:00:00 smflush
453 ? 00:00:00 spi0
507 ? 00:00:00 aoe_tx0
508 ? 00:00:00 aoe_ktio0
532 ? 00:00:00 raid5wq
534 ? 00:00:00 kmpathd
535 ? 00:00:00 kmpath_handlerd
568 ? 00:00:00 mv_crypto
598 ? 00:00:00 deferwq
604 ? 00:00:00 kworker/0:2
608 ? 00:00:00 kworker/u2:4
615 ? 00:00:00 kworker/u2:5
667 ? 00:00:00 kworker/0:1H
668 ? 00:00:00 kjournald
685 ? 00:00:00 kjournald
716 ? 00:00:00 buttonsd
728 ? 00:00:00 kjournald
758 ? 00:00:00 iscsi_eh
877 ? 00:00:00 fan-d
961 ? 00:00:00 kjournald
966 ? 00:00:00 kjournald
971 ? 00:00:00 udevd
1326 ? 00:00:00 httpd
1327 ? 00:00:01 httpd
1334 ? 00:00:00 noflushd
1337 ? 00:00:00 nmbd
1340 ? 00:00:00 smbd
1341 ? 00:00:00 smbd
1362 ? 00:00:00 syncawayd
1365 ? 00:00:00 discoverd
1373 ? 00:00:00 rpcbind
1389 ? 00:00:00 dropbear
1402 ? 00:00:00 hostd
1430 ? 00:00:00 spindown-usb-m
1433 ? 00:00:00 spindown-usb
1442 ? 00:00:00 temp-guard
1447 ? 00:00:00 ntpd
1475 ? 00:00:00 cmd_thread
1480 ? 00:00:00 wpa_supplicant
1489 ? 00:00:00 sh
1496 ? 00:00:00 httpd
1525 ? 00:00:00 kworker/0:3
1630 ? 00:00:00 dropbear
1631 ? 00:00:00 sh
1632 ? 00:00:00 sftp-server
1633 ? 00:00:00 dropbear
1634 ttyp0 00:00:00 sh
1650 ttyp0 00:00:00 ps

[Jocko]

Hi matzi,

I need another output with this command

Code: Select all

ps -ef |grep noflush


So please this output and detail what are the plugged disks

Note: usually, you don't need to enable the usb disk spin-down guard (service:spindown-usb-m)

[matzi]

Here it is:

root@fvdwsl-base:/ # ps -ef | grep noflush
root 1337 1 0 17:57 ? 00:00:00 /usr/sbin/noflushd -n 10 /dev/sda /dev/sdb
root 1552 1538 0 18:19 ttyp0 00:00:00 grep noflush

what can be wrong?

[Jocko]

So your settings is right and if sda can spin down I assume that it is a disk issue for sdb.

Could you try this command when sdb seems not to spin down

Code: Select all

 hdparm -y /dev/sdb

and see if the disk spin down.

You can play also with the flag S to set a spindown value; to restore vendor timeout

Code: Select all

hdparm -S 253 /dev/sdb

[matzi]

the disc in slot1 (it is sda I think) does an attempt to spin down but spins up again immediately. Some process must ask for it.

[Jocko]

Sorry, I don't know why I understood that your problem was the sdb disk :mrgreen:

In your output, there Are some not standard service

Code: Select all

453 ? 00:00:00 spi0
...
1430 ? 00:00:00 spindown-usb-m
1433 ? 00:00:00 spindown-usb
...
1442 ? 00:00:00 temp-guard
...
1475 ? 00:00:00 cmd_thread
1480 ? 00:00:00 wpa_supplicant
...
1632 ? 00:00:00 sftp-server


sftp-server is useless, also temp-guard with the fan control, stop the spindown usb guard (spindown-usb-m; spindown-usb) and spi0 & cmd_thread (bfgminer ???)

Look also if wpa_supplicant doesn't try to write periodically some data

[matzi]

I need to come back on this point (the actual soccer game gives me some time to think about different things).

Still my disc doesn't spin down. I stopped the spindown usb task and the temp guard. sftp can be open because I use winscp over sftp. So it will be down when I am disconnected.

I can kill cmd_thread but it will come back after reboot.
I can also kill spi but it comes back on spot.

But I found something more strange. See this:

root@fvdwsl-base:/ # noflushd -v
No devices given - autoprobing.
/dev/sda: using spindown handler SGIO
Error: No valid timeout found for /dev/sda
/dev/sdb: using spindown handler SGIO
Error: No valid timeout found for /dev/sdb

Does that mean there is a driver problem? So the daemon is set with correct values, it just does not know how to handle the disc? Can that be the reason hdparm does not succeed in spinning down?

Edit see the status after attempt to spin down:

root@fvdwsl-base:/ # hdparm -y /dev/sd?

/dev/sda:
issuing standby command

/dev/sdb:
issuing standby command
root@fvdwsl-base:/ # hdparm -C /dev/sd?

/dev/sda:
drive state is: active/idle

/dev/sdb:
drive state is: standby

[fvdw]

I have installed two discs, one in slot 1 one in slot 2. One or two of the discs is not spinning down even after hours

which one is n ot spinning down, the one in slot 1 or slot 2
the one is slot1 = sda the one is slot 2 is sdb

[fvdw]

using a command like "noflushd -v" is useless, it will give always output as you got

[fvdw]

I think wpa_supplicant might be the cause that prevent the disk form spinning down