DSM-G600, DNS-3xx and NSA-220 Hack Forum

perssinaasappel · 2008-11-11 18:04:22

After following the general tut for packages I managed to install smartmontools.

Code:

funpkg -i smartmontools-5.38-3.tgz

Using "Search" on this forum learned me to edit /ffp/etc/smartd.conf
The only thing I changed in here is:

Code:

#DEVICESCAN

/dev/sda -d marvell
/dev/sdb -d marvell

After starting the daemon /var/log/messages is filled with weird messages I don't understand.

Code:

Nov 11 14:40:00 CH3SNAS cron.notice crond[1591]: USER root pid 6763 cmd /usr/sbin/getdhcp&
Nov 11 14:41:11 CH3SNAS user.warn kernel: program smartd is using a deprecated SCSI ioctl, please convert it to SG_IO
Nov 11 14:41:12 CH3SNAS user.warn kernel: program smartd is using a deprecated SCSI ioctl, please convert it to SG_IO
Nov 11 14:41:12 CH3SNAS user.warn kernel: program smartd is using a deprecated SCSI ioctl, please convert it to SG_IO
Nov 11 14:41:12 CH3SNAS user.warn kernel: program smartd is using a deprecated SCSI ioctl, please convert it to SG_IO
Nov 11 14:41:12 CH3SNAS user.warn kernel: program smartd is using a deprecated SCSI ioctl, please convert it to SG_IO
Nov 11 14:41:12 CH3SNAS user.warn kernel: program smartd is using a deprecated SCSI ioctl, please convert it to SG_IO
Nov 11 14:41:12 CH3SNAS daemon.info smartd[6381]: Device: /dev/sdb, SMART Usage Attribute: 194 Temperature_Celsius changed from 118 to 119
Nov 11 14:41:12 CH3SNAS user.warn kernel: program smartd is using a deprecated SCSI ioctl, please convert it to SG_IO
Nov 11 14:41:12 CH3SNAS user.warn kernel: program smartd is using a deprecated SCSI ioctl, please convert it to SG_IO

What am I doing wrong here?

Can someone post his smartd.conf so I can see and learn from it?

Thx!

kirill · 2008-11-12 13:16:17

Try to specify -d ata instead of -d marvell

Last edited by kirill (2008-11-12 13:20:45)

perssinaasappel · 2008-11-12 13:44:56

Code:

root@CH3SNAS:~# smartctl -d ata /dev/sda
smartctl version 5.38 [arm-unknown-linux-uclibc] Copyright (C) 2002-8 Bruce Allen
Home page is http://smartmontools.sourceforge.net/

Smartctl: Device Read Identity Failed (not an ATA/ATAPI device)

A mandatory SMART command failed: exiting. To continue, add one or more '-T permissive' options.
root@CH3SNAS:~#

Thx for your input but only '-d marvell' is doing something.

[edit]Your first post! Welcome kirill.[/edit]

Last edited by perssinaasappel (2008-11-12 13:46:52)

bfg100k · 2009-05-22 15:55:11

I'm getting the warning about smartd using a deprecated SCSI ioctl too and its really filling up my log. Is there any way to get rid of it?

talkingRock · 2009-05-22 21:51:09

-d marvell is definitely what you want, not -d ata

I had this same problem and never found a direct solution. (See my workaround below.)
Presumably, there is a way of filtering using
klogd or syslogd which will do what we want but I did not find it.

For other reasons, I had a "dailyChores.sh" script that runs once per day.
I wrote a little section that removes the offending lines from the log.
In addition, rather than running the daemon full time, this script does a
a health check once per day, just before I clean the log.

perssinaasappel · 2009-05-24 00:20:43

For the same reason I stopped using the daemon and run a check every monday with smartctl.
The output is send to a file served by lighty, so I can see the results in my webbrowser.
It still generates the error in /var/log/messages but only once a week and I can live with that.

Added as a cronjob:

Code:

55 6 * * 1 /mnt/HD_a2/ffp/sbin/smartctl --attributes --device=marvell /dev/sda > /mnt/HD_a2/www/pages/smartmon/Volume_1.txt 2>&1
56 6 * * 1 /mnt/HD_a2/ffp/sbin/smartctl --attributes --device=marvell /dev/sdb > /mnt/HD_a2/www/pages/smartmon/Volume_2.txt 2>&1

bound4h · 2010-11-13 04:39:24

Fast forward 2 years from when this thread was started...does anyone know why this is happening? I'm trying to use messages as efficiently as possible but it fills up with these lines daily...

Pejn · 2010-11-13 15:00:54

Hmm im using smartmontools-5.38 with cronjob every 5min to log the temp on my hdds havent had any warnings what i know of

eastpole · 2011-04-15 19:39:21

I think this problem still exists. I have smartctl 5.40 from ipkg/optware and I see the messages:

Code:

program smartctl is using a deprecated SCSI ioctl, please convert it to SG_IO

It is in my dmesg buffer 202 times. I am logging the disk temp every 10 minutes, and not running the smart daemon.

Incidentally, my drives run ~30 C with the case fan at max, and approximately 10 C hotter with the case fan on automatic (mostly off) and to me that seems quite comfortable. The reason I want to log is because the box is soon to be exported to a distant friend's unventilated cupboard for my offsite backups. I'd hate for the drives to melt into pools of metal and mess up his cupboard.

smartctl version detail follows.

Cheers,
eastpole

Code:

root@exile:/mnt/HD_a2/bin# smartctl --version
smartctl 5.40 2010-10-16 r3189 [arm-unknown-linux-gnu] (local build)
Copyright (C) 2002-10 by Bruce Allen, http://smartmontools.sourceforge.net

smartctl comes with ABSOLUTELY NO WARRANTY. This is free
software, and you are welcome to redistribute it under
the terms of the GNU General Public License Version 2.
See http://www.gnu.org for further details.

smartmontools release 5.40 dated 2010-10-16 at 16:34:38 UTC
smartmontools SVN rev 3189 dated 2010-10-16 at 16:35:49
smartmontools build host: arm-unknown-linux-gnu
smartmontools build configured: 2010-11-02 03:22:08 UTC
smartctl compile dated Nov  2 2010 at 03:22:23
smartmontools configure arguments:  '--build=i386-pc-linux-gnu' '--host=arm-linux' '--target=arm-linux' '--prefix=/opt' '--disable-nls' '--disable-static' 'build_alias=i386-pc-linux-gnu' 'host_alias=arm-linux' 'target_alias=arm-linux' 'CXX=/home/slug/optware/dns323/toolchain/uclibc-toolchain-src-20040609/gcc-3.3.x/toolchain_arm/bin/arm-linux-uclibc-g++' 'LDFLAGS= -L/home/slug/optware/dns323/staging/opt/lib -Wl,-rpath,/opt/lib -Wl,-rpath-link,/home/slug/optware/dns323/staging/opt/lib ' 'CPPFLAGS=-O2   -pipe -fPIC -I/home/slug/optware/dns323/staging/opt/include ' 'CC=/home/slug/optware/dns323/toolchain/uclibc-toolchain-src-20040609/gcc-3.3.x/toolchain_arm/bin/arm-linux-uclibc-gcc'

Last edited by eastpole (2011-04-15 19:42:06)

karlrado · 2011-04-18 17:35:40

There is another thread on here someplace (sorry, I can't do the search for you now) about a guy who hacked the kernel to suppress the messages.

This solves the problem, but is fairly hard to do and I don't think it is the right approach.

I would bring up the issue with the smartmontools author(s), as the cleanest fix is for them to modify the sources to stop using the deprecated API's. Then hopefully, the optware folks would pick up the fix and release an update for ARM.

A lot needs to happen to get this fixed, but the first step would be to raise this with the smartmon people. I might look into this myself this week if I get a moment. I'd also like it fixed.

I suppose another approach would be to somehow deal with the messages after they are generated. Someone mentioned that the message log can cleaned up periodically, perhaps by another cron job.

FunFiler · 2011-04-18 18:30:19

As part of a nightly script, I dump and clear the dmesg log.

Code:

LOG_FILE="/mnt/usb/logs/dmesg.log"
echo "`date` Dumping dmesg information to ${LOG_FILE}"
echo "`date` System dmesg information dump" > ${LOG_FILE}
/ffp/bin/dmesg -c >> ${LOG_FILE}

It is easy enough to simply ignore the messages that are meaningless.

Last edited by FunFiler (2011-04-18 18:30:39)

carylewis2009 · 2011-04-20 18:40:54

I have taken a quick look at the os_linux.cpp code, and it appears as though the proper fix for this is to implement the SG_IO calls rather than the SCSI itocl calls currently being used in the Marvell classes, but I wonder if the Marvell hardware supports those calls.

Does anyone know?

DSM-G600, DNS-3xx and NSA-220 Hack Forum

Announcement

#1 2008-11-11 18:04:22

smartmontools

Code:

Code:

Code:

#2 2008-11-12 13:16:17

Re: smartmontools

#3 2008-11-12 13:44:56

Re: smartmontools

Code:

#4 2009-05-22 15:55:11

Re: smartmontools

#5 2009-05-22 21:51:09

Re: smartmontools

#6 2009-05-24 00:20:43

Re: smartmontools

Code:

#7 2010-11-13 04:39:24

Re: smartmontools

#8 2010-11-13 15:00:54

Re: smartmontools

#9 2011-04-15 19:39:21

Re: smartmontools

Code:

Code:

#10 2011-04-18 17:35:40

Re: smartmontools

#11 2011-04-18 18:30:19

Re: smartmontools

Code:

#12 2011-04-20 18:40:54

Re: smartmontools

Board footer