xfs
[Top] [All Lists]

Re: easily reproducible filesystem crash on rebuilding array

To: Christoph Hellwig <hch@xxxxxxxxxxxxx>
Subject: Re: easily reproducible filesystem crash on rebuilding array
From: Emmanuel Florac <eflorac@xxxxxxxxxxxxxx>
Date: Tue, 16 Dec 2014 12:38:58 +0100
Cc: Dave Chinner <david@xxxxxxxxxxxxx>, xfs@xxxxxxxxxxx
Delivered-to: xfs@xxxxxxxxxxx
In-reply-to: <20141216075605.GA24633@xxxxxxxxxxxxx>
Organization: Intellique
References: <20141211123936.1f3d713d@xxxxxxxxxxxxxxxxxxxx> <20141215130715.4dfaaa8e@xxxxxxxxxxxxxxxxxxxx> <20141215132500.13210fdb@xxxxxxxxxxxxxxxxxxxx> <20141215201036.GQ24183@dastard> <20141216075605.GA24633@xxxxxxxxxxxxx>
Le Mon, 15 Dec 2014 23:56:05 -0800
Christoph Hellwig <hch@xxxxxxxxxxxxx> Ãcrivait:

> On Tue, Dec 16, 2014 at 07:10:36AM +1100, Dave Chinner wrote:
> > The controller is broken if it's returning EIO to reads when it
> > is busy.
> 
> What controller is this? 

ASR-71685, but the problem occured several times with various ASR-7xx5
controllers and different firmware and drivers. 

> SCSI devices can return a QUEUE BUSY
> indicator, so having a RAID controller do something similar doesn't
> sound unusual.  But the driver needs to returns translate that into
> a QUEUE BUSY so that the SCSI midlayer can handle it correctly.
> 
> It might make sense to take this to linux-scsi with the driver
> maintainer in Cc.

The driver would be either aacraid or sd, then?

-- 
------------------------------------------------------------------------
Emmanuel Florac     |   Direction technique
                    |   Intellique
                    |   <eflorac@xxxxxxxxxxxxxx>
                    |   +33 1 78 94 84 02
------------------------------------------------------------------------

<Prev in Thread] Current Thread [Next in Thread>