On 12/29/2011 8:39 AM, Mauro Sánchez wrote:

> As I said before, in my opinion, there is a problem with the kernel
> and the disks. But it's just my opinion based on all the tests I've
> done.

The problem is with the kernels and modules you are using, and remotely
possibly teh 9240, not the disk drives.

Is BBWC enabled on this 9240? Is the cache is write-back mode? If so
disabled the cache or change its mode to write-through.

--
Stan


--
To UNSUBSCRIBE, email to debian-user-REQUEST@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.org
Archive: 4EFC869C.2070802@hardwarefreak.com">http://lists.debian.org/4EFC869C.2070802@hardwarefreak.com

2011/12/29 Camaleón <noelamac@gmail.com>:
> On Thu, 29 Dec 2011 10:31:28 -0300, Mauro Sánchez wrote:
>
>> Camaleon, the configuration with one disk uses JBOD. I have tried Debian
>> Stable with only one disk, RAID 1, RAID 5 and LVM. None of them worked.
>
> Did you try to partition and format the disks from Gparted?
>
Camaleon,
I created a partition of 10 GB with Gparted. Then I started to install
Squeeze but when is installing the base system it hangs up at 17% with
the following errors catched from the dmesg:
megasas: RESET -4637 cmd=2a retries=0
megasas: waiting for 1 commands to complete
megasas: failed to do reset
megasas: cannot recover form previous reset failures

And other errors of the I/O type.


--
To UNSUBSCRIBE, email to debian-user-REQUEST@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.org
Archive: CAMCF3yxGRrAnj+YKgNEu4TfD7Ln+Ef8Ug_fLr+We5an+7MOKg w@mail.gmail.com">http://lists.debian.org/CAMCF3yxGRrAnj+YKgNEu4TfD7Ln+Ef8Ug_fLr+We5an+7MOKg w@mail.gmail.com

2011/12/29 Stan Hoeppner <stan@hardwarefreak.com>:
> On 12/29/2011 8:39 AM, Mauro Sánchez wrote:
>
>> As I said before, in my opinion, there is a problem with the kernel
>> and the disks. But it's just my opinion based on all the tests I've
>> done.
>
> The problem is with the kernels and modules you are using, and remotely
> possibly teh 9240, not the disk drives.
>
> Is BBWC enabled on this 9240? *Is the cache is write-back mode? *If so
> disabled the cache or change its mode to write-through.
>
Hello Stan.
The disk cache is Disabled and de write mode is write through.
About BBWC I don't know where to check it.

Mauro.


--
To UNSUBSCRIBE, email to debian-user-REQUEST@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.org
Archive: CAMCF3ywXsM9z7xzwaWqQVsDBnPaVQ1F6vfUud0e_8wMqtYWs7 g@mail.gmail.com">http://lists.debian.org/CAMCF3ywXsM9z7xzwaWqQVsDBnPaVQ1F6vfUud0e_8wMqtYWs7 g@mail.gmail.com

On Thu, 29 Dec 2011 12:41:25 -0300, Mauro Sánchez wrote:

> 2011/12/29 Camaleón <noelamac@gmail.com>:
>> On Thu, 29 Dec 2011 10:31:28 -0300, Mauro Sánchez wrote:
>>
>>> Camaleon, the configuration with one disk uses JBOD. I have tried
>>> Debian Stable with only one disk, RAID 1, RAID 5 and LVM. None of them
>>> worked.
>>
>> Did you try to partition and format the disks from Gparted?
>>
> Camaleon,
> I created a partition of 10 GB with Gparted. Then I started to install
> Squeeze but when is installing the base system it hangs up at 17% with
> the following errors catched from the dmesg:
> megasas: RESET -4637 cmd=2a retries=0
> megasas: waiting for 1 commands to complete
> megasas: failed to do reset
> megasas: cannot recover form previous reset failures
>
> And other errors of the I/O type.

Ugh... that sounds like the RAID controller is hanging/resetting very badly
or the driver is behaving weirdly :-(

It can be then a bug in the megasas kernel module and Google finds *lots*
of similar errors for that same driver, and curiously, xen kernels:

http://www.google.com/webhp?complete=0&hl=en#hl=en&tbo=1&complete=0&site =webhp&tbs=qdr:y&q=LSI+megasas:+RESET+-4637+cmd%3D2a+retries%3D0&tbo=1&bav=on.2,or.r_gc.r _pw.,cf.osb&fp=e8f1b276e45aa3ef&biw=1280&bih=888

I wouldn't discard this is still happening on wheezy... although the
installer can succeed, what will happen once the system enters into
high load? Will the driver experience the same problems?

Are you using the NET ISO image to install Debian? If no, I would do a
final try using it, just to discard bad media and/or DVD read errors.

Greetings,

--
Camaleón


--
To UNSUBSCRIBE, email to debian-user-REQUEST@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.org
Archive: pan.2011.12.29.16.26.28@gmail.com">http://lists.debian.org/pan.2011.12.29.16.26.28@gmail.com

2011/12/29 Camaleón <noelamac@gmail.com>:
> On Thu, 29 Dec 2011 12:41:25 -0300, Mauro Sánchez wrote:
>
>> 2011/12/29 Camaleón <noelamac@gmail.com>:
>>> On Thu, 29 Dec 2011 10:31:28 -0300, Mauro Sánchez wrote:
>>>
>>>> Camaleon, the configuration with one disk uses JBOD. I have tried
>>>> Debian Stable with only one disk, RAID 1, RAID 5 and LVM. None of them
>>>> worked.
>>>
>>> Did you try to partition and format the disks from Gparted?
>>>
>> Camaleon,
>> I created a partition of 10 GB with Gparted. Then I started to install
>> Squeeze but when is installing the base system it hangs up at 17% with
>> the following errors catched from the dmesg:
>> megasas: RESET -4637 cmd=2a retries=0
>> megasas: waiting for 1 commands to complete
>> megasas: failed to do reset
>> megasas: cannot recover form previous reset failures
>>
>> And other errors of the I/O type.
>
> Ugh... that sounds like the RAID controller is hanging/resetting very badly
> or the driver is behaving weirdly :-(
>
> It can be then a bug in the megasas kernel module and Google finds *lots*
> of similar errors for that same driver, and curiously, xen kernels:
>
> http://www.google.com/webhp?complete=0&hl=en#hl=en&tbo=1&complete=0&site =webhp&tbs=qdr:y&q=LSI+megasas:+RESET+-4637+cmd%3D2a+retries%3D0&tbo=1&bav=on.2,or.r_gc.r _pw.,cf.osb&fp=e8f1b276e45aa3ef&biw=1280&bih=888
>
> I wouldn't discard this is still happening on wheezy... although the
> installer can succeed, what will happen once the system enters into
> high load? Will the driver experience the same problems?
>
Hello Camaleon.
I have installed debian testing and then installed xen-linux-system
and boot from its kernel. Till now everything just works fine. What
kind of tests would you suggest me for high load activity with the
controller?

> Are you using the NET ISO image to install Debian? If no, I would do a
> final try using it, just to discard bad media and/or DVD read errors.
>
I tried with the net iso and the iso cd, just to be sure it wasn't a
media error. Thanks for asking anyway.

> Greetings,
>
> --
> Camaleón
Greetings,
Mauro.


--
To UNSUBSCRIBE, email to debian-user-REQUEST@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.org
Archive: CAMCF3ywJvNdoiwK38nLWw=1+uo21W_PkQp3KgUvUhddiE29a= g@mail.gmail.com">http://lists.debian.org/CAMCF3ywJvNdoiwK38nLWw=1+uo21W_PkQp3KgUvUhddiE29a= g@mail.gmail.com

On Thu, 29 Dec 2011 13:39:18 -0300, Mauro Sánchez wrote:

> 2011/12/29 Camaleón <noelamac@gmail.com>:

(...)

>> It can be then a bug in the megasas kernel module and Google finds
>> *lots* of similar errors for that same driver, and curiously, xen
>> kernels:
>>
>> http://www.google.com/webhp?complete=0&hl=en#hl=en&tbo=1&complete=0&site =webhp&tbs=qdr:y&q=LSI+megasas:+RESET+-4637+cmd%3D2a+retries%3D0&tbo=1&bav=on.2,or.r_gc.r _pw.,cf.osb&fp=e8f1b276e45aa3ef&biw=1280&bih=888
>>
>> I wouldn't discard this is still happening on wheezy... although the
>> installer can succeed, what will happen once the system enters into
>> high load? Will the driver experience the same problems?
>>
> Hello Camaleon.
> I have installed debian testing and then installed xen-linux-system and
> boot from its kernel. Till now everything just works fine. What kind of
> tests would you suggest me for high load activity with the controller?

(...)

Compiling a kernel it usually puts a lot of stress into the system. If you
have the time and want to give it a whirl, it would be a good performance
load test :-)

Or just give the system a time and monitor how it behaves. I would
install the RAID controller utilities to get a closer look of the array
status just in the event it suddenly goes offline...

Greetings,

--
Camaleón


--
To UNSUBSCRIBE, email to debian-user-REQUEST@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.org
Archive: pan.2011.12.29.17.00.36@gmail.com">http://lists.debian.org/pan.2011.12.29.17.00.36@gmail.com

On 12/29/2011 9:49 AM, Mauro Sánchez wrote:
> 2011/12/29 Stan Hoeppner <stan@hardwarefreak.com>:
>> On 12/29/2011 8:39 AM, Mauro Sánchez wrote:
>>
>>> As I said before, in my opinion, there is a problem with the kernel
>>> and the disks. But it's just my opinion based on all the tests I've
>>> done.
>>
>> The problem is with the kernels and modules you are using, and remotely
>> possibly teh 9240, not the disk drives.
>>
>> Is BBWC enabled on this 9240? Is the cache is write-back mode? If so
>> disabled the cache or change its mode to write-through.
>>
> Hello Stan.
> The disk cache is Disabled and de write mode is write through.
> About BBWC I don't know where to check it.

Battery Backed Write Cache means you have the battery module installed
on the card, which enables write caching. If your mode was set to
write-through then you probably don't have the battery module. In any
case, if it's write-through, then the write cache isn't the problem.

--
Stan


--
To UNSUBSCRIBE, email to debian-user-REQUEST@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.org
Archive: 4EFD21DC.4040107@hardwarefreak.com">http://lists.debian.org/4EFD21DC.4040107@hardwarefreak.com

On 12/29/2011 9:41 AM, Mauro Sánchez wrote:
> 2011/12/29 Camaleón <noelamac@gmail.com>:
>> On Thu, 29 Dec 2011 10:31:28 -0300, Mauro Sánchez wrote:
>>
>>> Camaleon, the configuration with one disk uses JBOD. I have tried Debian
>>> Stable with only one disk, RAID 1, RAID 5 and LVM. None of them worked.
>>
>> Did you try to partition and format the disks from Gparted?
>>
> Camaleon,
> I created a partition of 10 GB with Gparted. Then I started to install
> Squeeze but when is installing the base system it hangs up at 17% with
> the following errors catched from the dmesg:
> megasas: RESET -4637 cmd=2a retries=0
> megasas: waiting for 1 commands to complete
> megasas: failed to do reset
> megasas: cannot recover form previous reset failures
>
> And other errors of the I/O type.

This is all brand new hardware isn't it? I've been working under that
assumption. If it is used, or refurb stuff, scratch my previous
statement about the drives, as they could very well have problems, or
anything else in the chassis for that matter.

New/used/refurb, the first thing I would do, given those errors, is
check/replace the 8087 cable connecting the 9240 to the drive cage
backplane.

--
Stan



--
To UNSUBSCRIBE, email to debian-user-REQUEST@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.org
Archive: 4EFD23B0.2060708@hardwarefreak.com">http://lists.debian.org/4EFD23B0.2060708@hardwarefreak.com

On 12/29/2011 10:39 AM, Mauro Sánchez wrote:

> I have installed debian testing and then installed xen-linux-system
> and boot from its kernel. Till now everything just works fine. What
> kind of tests would you suggest me for high load activity with the
> controller?

Bonnie++, IOzone, FIO, etc. Anything with a high random IOPS workload.
Just running a huge sequential write/read with dd probably won't cause
any problems to surface.

--
Stan



--
To UNSUBSCRIBE, email to debian-user-REQUEST@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.org
Archive: 4EFD24F3.70309@hardwarefreak.com">http://lists.debian.org/4EFD24F3.70309@hardwarefreak.com

2011/12/29 Stan Hoeppner <stan@hardwarefreak.com>:
> On 12/29/2011 9:41 AM, Mauro Sánchez wrote:
>> 2011/12/29 Camaleón <noelamac@gmail.com>:
>>> On Thu, 29 Dec 2011 10:31:28 -0300, Mauro Sánchez wrote:
>>>
>>>> Camaleon, the configuration with one disk uses JBOD. I have tried Debian
>>>> Stable with only one disk, RAID 1, RAID 5 and LVM. None of them worked.
>>>
>>> Did you try to partition and format the disks from Gparted?
>>>
>> Camaleon,
>> I created a partition of 10 GB with Gparted. Then I started to install
>> Squeeze but when is installing the base system it hangs up at 17% with
>> the following errors catched from the dmesg:
>> megasas: RESET -4637 cmd=2a retries=0
>> megasas: waiting for 1 commands to complete
>> megasas: failed to do reset
>> megasas: cannot recover form previous reset failures
>>
>> And other errors of the I/O type.
>
> This is all brand new hardware isn't it? *I've been working under that
> assumption. *If it is used, or refurb stuff, scratch my previous
> statement about the drives, as they could very well have problems, or
> anything else in the chassis for that matter.
>
> New/used/refurb, the first thing I would do, given those errors, is
> check/replace the 8087 cable connecting the 9240 to the drive cage
> backplane.
>
Hello Stan, it's all brand new hardware. So I think it's not a
hardware problem, because I have installed Wheezy on it without any
problem.


--
To UNSUBSCRIBE, email to debian-user-REQUEST@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.org
Archive: CAMCF3yxB9BzrUKtoH4KEymJu1B_8s8RrNgWAaajXUNM+F3czz Q@mail.gmail.com">http://lists.debian.org/CAMCF3yxB9BzrUKtoH4KEymJu1B_8s8RrNgWAaajXUNM+F3czz Q@mail.gmail.com