FreeBSD 3ware & Maxtor problems

[Philip Jenvey]

Carroll -

Sure, I've cced -stable.

I am actually using the card on a 66mhz bus (no riser card), and I just 
noticed the Knowledge Base article before you mailed me back. The PDF 
link you posted is broken.

Is an RMA necessary in my case? It appears they are providing a 
firmware upgrade that's supposed to take care of the problem (and my 
firmware is older). Installing that is my next step.


On Sep 28, 2004, at 4:36 PM, Carroll Kong wrote:

> Is it okay if I cc this to the mailing lists?  I think a lot of people 
> might
> benefit from this.
>
> ----- Original Message -----
> From: "Philip Jenvey" <pjenvey at groovie.org>
> To: <me at carrollkong.com>
> Sent: Tuesday, September 28, 2004 6:52 PM
> Subject: FreeBSD 3ware & Maxtor problems
>
>
>> Carroll -
>>
>> I noticed your postings re your RAID with Maxtor disks.
>>
>> I'm having similar issues, with a 8506-4LP + 4 Maxtor DiamondMax 9
>> 160GBs. Basically I've been getting some kind of drive error that
>> typically leads to the system hanging, and causes the RAID to be
>> degraded upon reboot. Most of the time I've been able to rebuild and
>> continue on.
>
> If you are using it in a 66 mhz slot, that is probably your problem 
> right
> there.  If you are using riser cards, and the controller card was 
> purchased
> before Q1 2004, then you need to buy a new riser card.  There is a 
> physical
> bug with the card handling 66 mhz properly.  Even if you directly 
> connect to
> the system, you can run into issues if the system mobo supports 66 
> mhz.  I
> believe they will accommodate RMAs in this case.
>
> https://www.3ware.com/kbadmin/attachments/T...20Rev%20A_P.pdf
> http://www.3ware.com/KB/article.aspx?id=10964
>
> Now, if none of that appiles to you, then the problem continues
>
> http://forums.storagereview.net/index.php?showtopic=14162
>
> Scroll to the last message.
>
> Implicitly they are saying that a lot of CRC errors can cause the 
> drive to
> "drop a drive" from the array.  In my case, I am using an older ATA
> backplane which only supports ATA/33... I am suspecting the CRC errors 
> are
> related to my backplane.  I am getting it swapped out tomorrow, so 
> hopefully
> that will resolve that issue.  If not, I have to keep replacing 
> parts...
> such as my controller card itself and/or the riser card (even though I 
> am
> not affected by the 66 mhz issue).  I do get random harddisk dropping 
> out of
> my array.  I am suspecting that if my controller gets one of these 
> major
> drop outs during operation, it probably can result in a hard hang.  Bad
> cables can cause this problem as well.
>
> Random loss of power to the disk might cause it to drop out of the 
> array as
> well.
>
>> What's the latest you've found on this issue?
>>
>> I noticed you were playing with the smart monitoring tools, have they
>> shed any light on the subject? I'm planning to try them but this RAID
>> has been down for about a half day now (at this point I'm not sure if 
>> a
>> drive is dead or what, I seem to be getting errors on 2 different
>> drives now).
>
> Smartctl helped me pre-emptively remove some disks that was failing,
> however, that was not the problem.  Even after removing one bad disk 
> and
> rebuilding the array, I still ran into the same problems.
>
>> Have you contacted 3ware at all about the problem?
>
> I have recently contacted 3ware by emailing the results from the 3Ware 
> cli
> program.  If your 3dmd or 3dmd2 supports it, go to the 'alarms' 
> section and
> download the extended errorlog file and send it off to them.  If you 
> do not
> have this feature or all you see is /var/log/messages kind of output 
> then do
> this.
>
> (do something like)
> ./tw_cli
> info c0 errorlog
>
> and copy that output to a file.  Also send them a copy of 3dmd's 
> technical
> page.
>
> Write down some simple steps on what you did and how to repeat the 
> problem
> if possible.
>
> Then email it off to
> support at webrt.3ware.com
>
>> Thanks,
>>
>> --
>> Philip Jenvey
>
>
>
> - Carroll Kong
>
>
>
--
Philip Jenvey