Re: Areca RAID Controler and commit 98dcc2946adbe4349ef1ef9b99873b912831edd4

["Christian Bahls - gmail.com" <christian.bahls@gmail.com>]

Hello James

I used the two most recent debian kernels (3.10 and 3.11)
both show this behaviour, both have the same problem

Also of course i had a look at both git trees before sending out the email

Taking the debian kernel source tree for 3.11-trunk and reverting the patch
makes the raid controler work again

So this really isn't a fixed problem

I believe it should be fixed in 3.12 3.11 and any stable kernel tree around
(1 & 1 must have thousands of customers that might run into the problem in short time)

Kind regards
 Christian Bahls



[2013-09-22 19:05] James Bottomley <jbottomley@parallels.com> wrote:
> [cc to linux-scsi added]
> On Sun, 2013-09-22 at 16:27 +0200, Christian Bahls - gmail.com wrote:
> > Dear all,
> > 
> > Looks like i have been bitten by the above mentioned commit.
> 
> Just from a practical point of view, I only really work upstream, so
> this commit id doesn't show anything for me. I think the commit you're
> referring to is:
> 
> commit 66c28f97120e8a621afd5aa7a31c4b85c547d33d
> Author: Martin K. Petersen <martin.petersen@oracle.com>
> Date:   Thu Jun 6 22:15:55 2013 -0400
> 
>     [SCSI] sd: Update WRITE SAME heuristics
> 
> > I have several root server at German Hoster 1 & 1 (United Internet)
> > 
> > I have been updating my Kernel to version 3.10 recently - using
> > Debians update mechanism -and loosing the server afterwards
> > 
> > As lots of people will have this hardware -at leas in that datacenter-
> > will the fix mentioned in https://lwn.net/Articles/561356/ be hitting
> > linux-stable anytime soon?
> 
> It's been there for months.  It went to stable at the end of June.  If I
> look at the tree, it was in 3.10.3.  Debian testing is already on
> 3.10.7 ... not sure what the rest has because I only run testing.
> 
> James
> 

--