Re: [Qemu-devel] [PATCH] Asynchronous reliable and configurable cache flush

[Ian Jackson]

Paul Brook writes ("Re: [Qemu-devel] [PATCH] Asynchronous reliable and 
configurable cache flush"):
> [Ian Jackson:]
> > I don't think flush failures are a normal IO error.  They can only
> > occur when the corresponding host block device is having serious
> > trouble.  Normally (if it's actually a host disk) this is very bad.
> 
> In practice any IO error indicates that the target device is pretty screwed.
> I don't see why flushes should be special. When caches are enabled I'd be 
> fairly surprised if normal write operations ever returned an error.

The most common kind of `IO error' will be disk full (or the
equivalent) on the difference file for a COW device.  That can't be
detected at flush time (at least, not with most host filesystems).

Ian.