Re: sata_sil24 test support

Re: sata_sil24 test support

[James O. Rose, III]

On Wednesday 16 November 2005 11:03 pm, you wrote:

I'm keeping this mail intact for reference on linux-ide.

> James Rose wrote:
> > Mr. Heo:
> > 
> > I have a couple of questions regarding the sata_sil24 driver, and 
I would
> > also like to offer up some test support.
> > 
> > I'm by no means any kind of kernel hacker and only have a 
rudimentary
> > understanding of C at best, so I can't provide any code help.  I 
am
> > interested in running whatever tests you think could help you with 
the
> > driver development.  This box is basically not doing anything, so 
I can
> > run automated testing for days at a time.
> > 
> > Here's my h/w specs:
> > --------------------
> > Intel SE7620AF2 Motherboard
> > Dual XEON 3.2 GHz with HyperThreading
> > 4 Gig Memory
> > Silicon Image 3124
> > Silicon Image 3132
> > Rancho Technologies MiniStor 3 (Silicon Image 3726 Port Multiplier 
Chip)
> > with 4 100GB Hitachi Travelstar SATA Drives
> > SteelVine 2000 (Silicon Image 3726 Port Port Multiplier Chip) with 
5 400GB
> > Hitachi SATA Drives
> > Various External 500G SATA2 Drives
> > 
> > NOTE: All the hard drives are essentially blank, and can be set up 
as
> > RAID/LVM/Whatever, on a per test basis.
> > 
> > Here's my s/w specs:
> > --------------------
> > kernel 2.6.15rc1-netdev
> > Debian Sarge for x86 (with udev v0.056-3) [I'm working on getting 
64 bit
> > working as well, I'm just not there yet].
> > 
> > I regret I'm unable to provide any kind of shell access (corporate
> > firewall), but I will happily run whatever you think can help.  
I'm very
> > interested in seeing a fully functional (if it already is, my 
apologies. 
> > I need to figure out some stuff still) driver released into the 
kernel.
> > 
> > I'll happily provide lspci's, .configs, etc. upon request.
> 
> One thing I wanna verify on sil24 is data integrity with multiple 
disks 
> attached.  It would be very helpful if you can do some parallel data 
> stress testing with multiple disks.
> 
> * Parallel 'badblocks -w -t random' on all attached disks.  Maybe 
repeat 
> it for a few days and verify no corrupted IO occurs.
> 
> * Create a random 2G or 4G file on a disk (dd if=/dev/urandom 
> of=testfile bs=1M count=4K) and parallely fill all disks with the 
file 
> (for ((i=0;i<20;i++)); do cp testfile /mnt/sdx/testfile$i; done).  
After 
> disks are filled, calculate md5sum of all copies and verify that 
they 
> haven't corrupted.  Repeat for a few days.
> 

Thanks for getting back to me on this.  I'll try to set something up 
tomorrow.  BTW, is there any particular filesystem you'd prefer I 
test on (at least initially)?  I've currently got ext3 compiled in, 
but I can reconfigure to whatever.  I'll send results as I get them.

> > 
> > One Problem:
> > ------------
> > I am able to see the first drive on each of the port multipliers, 
but I
> > cannot seem to find a way to locate the other drives.  Is there an 
option
> > I can pass through modprobe, or some other kind of magic?
> 
> Unfortunately, port multiplier support is not implemented yet.  
Edward 
> Falk (CC'd above, Hi! Edward) used to work on PM support for sil24 a 
> while ago and might be able to help you out.
> 

Depressing news.  I'm willing to do what I can to help someone out 
with this.

> > If I should have sent this to someone else, please forward this 
along.  I
> > didn't know if this was appropriate for me to send to the LKML, so 
I sent
> > it to you instead (apologies if I was mistaken).  Please let me 
know what
> > I can (or can't) do.
> 
> If you have certain requests/questions regarding SATA in general or 
> specific SATA driver in particular, write a mail TO the maintainer 
and 
> CC linux-ide@vger.kernel.org.  And your mail is completely 
appropriate 
> for linux-ide.
> 

will do

> -- 
> tejun
> 
Regards,

James Rose

Re: sata_sil24 test support

[Tejun Heo]

James O. Rose, III wrote:
> 
> Thanks for getting back to me on this.  I'll try to set something up 
> tomorrow.  BTW, is there any particular filesystem you'd prefer I 
> test on (at least initially)?  I've currently got ext3 compiled in, 
> but I can reconfigure to whatever.  I'll send results as I get them.
> 

Thank you.  :-)  I'll be expecting to hear from you.

-- 
tejun

Re: sata_sil24 test support

[James Rose]

> James O. Rose, III wrote:
>>
>> Thanks for getting back to me on this.  I'll try to set something up
>> tomorrow.  BTW, is there any particular filesystem you'd prefer I
>> test on (at least initially)?  I've currently got ext3 compiled in,
>> but I can reconfigure to whatever.  I'll send results as I get them.
>>
>
> Thank you.  :-)  I'll be expecting to hear from you.
>
> --
> tejun
>

Results for first test (other badblock runs still going):

/dev/sdd is 100GB Hitachi Travelstar in a Rancho Technologies MiniStor 3
(Sil 3726), connected to internal SATA connector on Sil 3132

Hope this helps.  I'll continue to post results. Let me know if you need
more information, or if this is enough.

James
-------------------------------------
debian:~# time badblocks -w -t random /dev/sdd > badblocks.sdd.txt

real    106m1.969s
user    2m49.635s
sys     1m6.408s

debian:~# cat badblocks.sdd.txt
debian:~#

debian:~# fdisk -l /dev/sdd

Disk /dev/sdd: 100.0 GB, 100030242816 bytes
255 heads, 63 sectors/track, 12161 cylinders
Units = cylinders of 16065 * 512 = 8225280 bytes

Disk /dev/sdd doesn't contain a valid partition table
-------------------------------------------------------

Re: sata_sil24 test support

[James Rose]

Results for second test (one more badblock run still going):

/dev/sdd is 400GB Hitachi Deskstar in a SteelVine 2000
(Sil 3726), connected to external SATA connector on Sil 3124

------------------
debian:~# time badblocks -w -t random /dev/sdb > badblocks.sdb.txt

real    287m48.162s
user    11m49.856s
sys     4m14.692s
debian:~# cat badblocks.sdb.txt
debian:~# fdisk -l /dev/sdb

Disk /dev/sdb: 400.0 GB, 400088457216 bytes
255 heads, 63 sectors/track, 48641 cylinders
Units = cylinders of 16065 * 512 = 8225280 bytes

Disk /dev/sdb doesn't contain a valid partition table
debian:~#
---------------

Re: sata_sil24 test support

[James Rose]

Last one of the badblocks scans for Tejun:

dev/sdd is external 500GB SATA2 Hitachi Deskstar connected to Sil 3124

----------------------------------
debian:~# time badblocks -w -t random /dev/sdc > badblocks.sdc.txt

real    338m50.894s
user    14m41.059s
sys     5m15.016s
debian:~# cat badblocks.sdc.txt
debian:~# fdisk -l /dev/sdc

Disk /dev/sdc: 500.1 GB, 500107862016 bytes
255 heads, 63 sectors/track, 60801 cylinders
Units = cylinders of 16065 * 512 = 8225280 bytes

Disk /dev/sdc doesn't contain a valid partition table
debian:~#
----------------------------------

Hopefully tomorrow I can write a script for the write testing, so it can
run over the weekend.

Regards,

James

sata_sil24 corruption FIXED by motherboard swap

[linux]

For anybody who's been following my saga, what I thought was software
problems turned out to be a hardware problem with one PCI-express slot
on an Albatron K8NF4U motherboard.  (Whose support folks still haven't
gotten back to me...)

I swapped in a new motherboard this morning, and have been running disk
tests all afternoon with no problems.

My sincere apologies for the false alarm.

Re: sata_sil24 test support

[linux]

> One thing I wanna verify on sil24 is data integrity with multiple disks 
> attached.  It would be very helpful if you can do some parallel data 
> stress testing with multiple disks.
> 
> * Parallel 'badblocks -w -t random' on all attached disks.  Maybe repeat 
> it for a few days and verify no corrupted IO occurs.

I only ran it for a day, but I can report success on exactly this
test on 6x Seagate 7200.8 drives (350G partition of 400G drives)
across 3x Sil3132.

That's how I found my problems, and how I verified that they were gone.

(The only "intereting" finding was that one drive was noticeably slower
than the others.  Not 10%, but it finished most of an hour later.  I
checked the cables and all looked well, and its partner on the same
controller was fine.  I'm going to do a bit of swapping to experiment.)

This is with CONFIG_PCI_MSI=y.  It was run in single-user mode (all
file systems mounted read-only) because the question was whether
live file systems were safe.

One thing I'm thinking of as a *driver* test is to write a little utility
that uses O_DIRECT to do heavy I/O to the drive's cache.  That should
be able to exceed the 60 MB/sec media transfer rate limit.

Re: sata_sil24 test support

[linux]

> One thing I wanna verify on sil24 is data integrity with multiple disks 
> attached.  It would be very helpful if you can do some parallel data 
> stress testing with multiple disks.
> 
> * Parallel 'badblocks -w -t random' on all attached disks.  Maybe repeat 
> it for a few days and verify no corrupted IO occurs.

Just completed 6 passes x 6 drives x 350 GB = 12.6 TB of badblocks (10^14
bits) with no errors.  That's in addition to a previous 5 passes that
was interrupted by timeout problems on one drive, but that's an error
handling issue and not a data corruption problem, and it did resolve
itself eventually after I killed the badblocks run.

That's several days of solid disk access at > 300 MB/sec.
(Some silly people asked me why I ingored the Sil3114 that came with
the motherboard...)

Thanks for a great driver!  I'll have even more fun testing NCQ one of
these days. :-)

Now rebooting to 2.6.14-rc2.  Now that it's stable, this system is going
into production Very Very Soon.  If you want any more testing, speak up!

Re: sata_sil24 test support

[Tejun Heo]

linux@horizon.com wrote:
>>One thing I wanna verify on sil24 is data integrity with multiple disks 
>>attached.  It would be very helpful if you can do some parallel data 
>>stress testing with multiple disks.
>>
>>* Parallel 'badblocks -w -t random' on all attached disks.  Maybe repeat 
>>it for a few days and verify no corrupted IO occurs.
> 
> 
> Just completed 6 passes x 6 drives x 350 GB = 12.6 TB of badblocks (10^14
> bits) with no errors.  That's in addition to a previous 5 passes that
> was interrupted by timeout problems on one drive, but that's an error
> handling issue and not a data corruption problem, and it did resolve
> itself eventually after I killed the badblocks run.
> 
> That's several days of solid disk access at > 300 MB/sec.
> (Some silly people asked me why I ingored the Sil3114 that came with
> the motherboard...)
> 
> Thanks for a great driver!  I'll have even more fun testing NCQ one of
> these days. :-)
> 
> Now rebooting to 2.6.14-rc2.  Now that it's stable, this system is going
> into production Very Very Soon.  If you want any more testing, speak up!

I'm very glad to here the good news.  I'll let you know when more
testing is needed.  Thanks for doing this.

-- 
tejun