I have tried to find information about this question in google, but
found nothing.

The question is: are there issues using different disk i/o schedulers
(by appending elevator=xxx to the kernel boot command line) when using
md RAID1 or RAID5?

I seem to remember that someone told me that using elevator=deadline can
cause perfectly good disks to be kicked from the RAID because of some
latency issues that make the disk seem unresponsive (and dead) to the
raid software.

Is this true? Are there compatibility issues using different i/o
schedulers with software raid?


Thanks

--

Fabio "Kurgan" Muzzi

(worried Linux sysadmin)

--
To unsubscribe from this list: send the line "unsubscribe linux-raid" in
the body of a message to majordomo [at] vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html

On Thu, 29 Jul 2010, Fabio Muzzi wrote:

> Is this true? Are there compatibility issues using different i/o
> schedulers with software raid?

I'd actually like to raise this one level further:

In the case of (drives)->md->dm(crypto)->lvm->fs, how do the schedulers,
readahead settings, blocksizes, barriers etc interact thru all these
layers? Is block caching done on all layers? Is readahead done on all
layers?

--
Mikael Abrahamsson email: swmike [at] swm.pp.se
--
To unsubscribe from this list: send the line "unsubscribe linux-raid" in
the body of a message to majordomo [at] vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html

Fabio Muzzi wrote:
>
> I have tried to find information about this question in google, but
> found nothing.
>
> The question is: are there issues using different disk i/o schedulers
> (by appending elevator=xxx to the kernel boot command line) when using
> md RAID1 or RAID5?
>
> I seem to remember that someone told me that using elevator=deadline can
> cause perfectly good disks to be kicked from the RAID because of some
> latency issues that make the disk seem unresponsive (and dead) to the
> raid software.
>
> Is this true? Are there compatibility issues using different i/o
> schedulers with software raid?
>
>
> Thanks
>

I've been running several vmware server hosts with elevator=deadline on
raid1 for a year or so now, with no problems.

--
-Eric 'shubes'

--
To unsubscribe from this list: send the line "unsubscribe linux-raid" in
the body of a message to majordomo [at] vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html

On 29/07/10 19:23, Eric Shubert wrote:

> I've been running several vmware server hosts with elevator=deadline on
> raid1 for a year or so now, with no problems.

Nice to hear it, because I have just switched my newly installed vmware
server (Debian Lenny, md raid1, vmware server) from CFQ to deadline
because CFQ is VERY slow, causes the entire system to hang for some
seconds every now and then, and it seems that it also crashed the whole
machine during intense I/O from the only virtual guest I have running (a
Win2008 with MS SQL server).


--

Fabio "Kurgan" Muzzi

--
To unsubscribe from this list: send the line "unsubscribe linux-raid" in
the body of a message to majordomo [at] vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html