D2Plug - Log files flooded with error mesages related to /dev/sda

[kstaniek]

Hi,

I noticed that the log files: kern.log ; messages; syslog ; dmesg are flooded with continuously repeating error messages:

Code:

Nov 12 22:40:19 D2Plug kernel: [ 1312.394633] sd 1:0:0:0: [sda] Unit Not Ready
Nov 12 22:40:19 D2Plug kernel: [ 1312.394649] sd 1:0:0:0: [sda] Sense Key : 0x3 [current]
Nov 12 22:40:19 D2Plug kernel: [ 1312.394665] sd 1:0:0:0: [sda] ASC=0x30 ASCQ=0x0
Nov 12 22:40:19 D2Plug kernel: [ 1312.411258] sd 1:0:0:0: [sda] Unit Not Ready
Nov 12 22:40:19 D2Plug kernel: [ 1312.411271] sd 1:0:0:0: [sda] Sense Key : 0x3 [current]
Nov 12 22:40:19 D2Plug kernel: [ 1312.411286] sd 1:0:0:0: [sda] ASC=0x30 ASCQ=0x0
Nov 12 22:40:19 D2Plug kernel: [ 1312.415630] sd 1:0:0:0: ioctl_internal_command return code = 8000002
Nov 12 22:40:19 D2Plug kernel: [ 1312.415643]    : Sense Key : 0x3 [current]
Nov 12 22:40:19 D2Plug kernel: [ 1312.415655]    : ASC=0x30 ASCQ=0x0
Nov 12 22:40:19 D2Plug kernel: [ 1312.429134] sd 1:0:0:0: [sda] Unit Not Ready
Nov 12 22:40:19 D2Plug kernel: [ 1312.429146] sd 1:0:0:0: [sda] Sense Key : 0x3 [current]
Nov 12 22:40:19 D2Plug kernel: [ 1312.429161] sd 1:0:0:0: [sda] ASC=0x30 ASCQ=0x0
Nov 12 22:40:19 D2Plug kernel: [ 1312.433406] sd 1:0:0:0: ioctl_internal_command return code = 8000002
Nov 12 22:40:19 D2Plug kernel: [ 1312.433418]    : Sense Key : 0x3 [current]
Nov 12 22:40:19 D2Plug kernel: [ 1312.433430]    : ASC=0x30 ASCQ=0x0

The logfiles grow rapidly and daily size is about hundreds of megs. I guess the file system is heavy utilized because of this.

Anyone can give an advice or help in troubleshooting.
Cheers,
klaus

[mentor]

I assume you connect to usb flash-disk or harddisk through e-Sata

Do you mean these device not working or has some problem

if not , I think it just a information from kernel

not relate to your file-system

[kstaniek]

Nothing is connected to eSATA. I don't even have any eSATA device to check what happens when I connect it.

I don't think this is normal situation having a massive error message flooding to the log files.
It stopes when I ordinary remove the sda devce with 'rm /dev/sda'.
I don't think this is a solution. BTW I put this command into the /etc/rc.local.

It seems to be a bug in kernel.

Is anyone else facing the same issue?

[cbxbiker61]

Nothing is connected to eSATA. I don't even have any eSATA device to check what happens when I connect it.

I don't think this is normal situation having a massive error message flooding to the log files.
It stopes when I ordinary remove the sda devce with 'rm /dev/sda'.
I don't think this is a solution. BTW I put this command into the /etc/rc.local.

It seems to be a bug in kernel.

Is anyone else facing the same issue?

Install smartmontools if you don't have it already and run a "long" test.  That should tell you if you have a drive problem.

[kstaniek]

What for? I don't have drive connected to eSATA interface on D2PLUG.

On D2Plug:
/dev/sda seems to be eSATA - I have no eSATA to check it
/dev/sdb is SD Card (external)
/deb/sdc is USB drive

It look like kernel is trying to do something with device /dev/sda which generates this massive error log.

I don't see any reason for installing smartmontools to solve this issue.

[cbxbiker61]

Yeah, your right.  I didn't read your post carefully enough to see you don't have anything connected to /dev/sda.

I'd say you probably have some software that's checking for available devices periodically and causing the kernel to spit out the errors when it detects no device attached at sda.  If it really is the driver software polling the device by itself (without a software request), someone should probably fix that.

You could compile a custom kernel without the sata drivers included, that would eliminate it.

[mentor]

But I still suggest ,you could connect with USB-flash ,SD card , or hard-disk through eSATA to test it

this way could know even there are flooding message as you see ,the real function work normal or not

[kstaniek]

I've found that udisks-daemon was polling all the devs: /dev/sda /dev/sdb and /dev/sdc probably for automounting.

I just eliminated it by adding the following line:

Code:

KERNEL=="sda*", ENV{UDISKS_DISABLE_POLLING}="1"

to /lib/udev/rules.d/80-udisks.rules

however there's still open question remaining.
Is it a bug in kernel sata module or udisks itself causing error message when accessing not connected disk or just my controller port is broken.
Unfortunately I can't check it out because I haven't got the eSATA device.

Anyone can check the logs if the same error messages appear?

[vleek]

My D2Plug's /var/log/kern.log is also flooded, the same messages, size ~24MB
Thank you for the tip how to stop it.

[llamich]

I'm new in forum, and I have the same problem.
Well, I have problems to activate the bluetooth too. Due to the problem of the kernel log, I can not see what is happening with the bluetooth.

Does anyone has solution, though the problem?

Is tere a Globalscale support...?

Thank You

[kstaniek]

llamich: Try this:
http://www.plugcomputer.org/plugforum/index.php?topic=5942.0
and please avoid of using this thread to solve another problem. This is related to the log flooding issue.

So of there are others facing the same issue with log flooding it seems that this is not hw related issue.
Anyone from Global Scale/Marvell can have a look on this?
Cheers,
klaus

[Hell-G]

I have the same issue with log flooding. I noticed it at the very beginning, when I got the D2plug. So I guess it is already present in the original configuration.

I now plugged in my eSATA drive. It shows up in /var/log/messages but you have to look for it, as dmesg will be flooded again instantly. According to

Code:

fdisk -l

the eSATA drive is on /dev/sdc and not sda which is mentioned in dmesg.

I then plugged in another external drive on the USB port: it appears as /dev/sdd and again it is instantly hidden in dmesg. When I remove the eSATA drive and plug in the USB hard disk on the combined eSATA/USB port, it is recognised as /dev/sdc. Same thing happens when the drive is plugged in at boot.

So I don't think a drive on the eSATA port will solve this issue.

Has anyone tested what happens, when you insert a memory card?