Date: Wed, 8 Jul 2015 08:35:15 +0100 From: Steven Hartland <killing@multiplay.co.uk> To: Yamagi Burmeister <lists@yamagi.org> Cc: freebsd-scsi@freebsd.org Subject: Re: Device timeouts(?) with LSI SAS3008 on mpr(4) Message-ID: <559CD2B3.7000404@multiplay.co.uk> In-Reply-To: <20150708074652.07a815e6aa08526d569f3077@yamagi.org> References: <20150707132416.71b44c90f7f4cd6014a304b2@yamagi.org> <9426ced85d7def424e106fdefd7448ae@mail.gmail.com> <20150707183135.2c3f5aa45696b55a17e2f87f@yamagi.org> <559C0184.4050102@multiplay.co.uk> <20150708074652.07a815e6aa08526d569f3077@yamagi.org>
next in thread | previous in thread | raw e-mail | index | archive | help
Actually not, it could indicate a design problem with the midplane /
backplane is the cause of the issue.
We've had a number of Supermicro and Dell chassis when used in
combination with 6Gbps+ devices particularly SSD's that exhibit timeouts
like you describe, all turned out to be a backplane issue.
We proved this in by connecting the drives direct to the controller with
high quality cables eliminating the hotswap backplane, after which the
timeouts stopped.
This is a PITA to test as power is supplied by the hotswap backplane,
but I wouldn't recommend you look anywhere else till you've eliminated
this as a potential cause.
Regards
Steve
On 08/07/2015 06:46, Yamagi Burmeister wrote:
> Hello Steven,
> since the issue occures on all 3 servers it's at least unlikely. But
> I'll see what I can do.
>
> Regards,
> Yamagi
>
> On Tue, 7 Jul 2015 17:42:44 +0100
> Steven Hartland <killing@multiplay.co.uk> wrote:
>
>> Have you eliminated the midplane / cabling as the issue as that's very
>> common.
>>
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?559CD2B3.7000404>