[dm-devel] awful request merge results while simulating high IOPS multipath

[Busch, Keith]

Sorry, my reply was non sequitur to this thread. We don't do merging in NVMe.

Our first bottleneck appears to be the device mapper's single lock request queue.

> -----Original Message-----
> From: Jens Axboe [mailto:axboe at kernel.dk]
> Sent: Wednesday, February 25, 2015 8:12 AM
> To: Busch, Keith; Mike Snitzer
> Cc: jmoyer at redhat.com; dm-devel at redhat.com; shivakrishna.merla at netapp.com; Bart Van Assche
> Subject: Re: awful request merge results while simulating high IOPS multipath
> 
> On 02/24/2015 08:14 PM, Keith Busch wrote:
> > To be honest, we also see underwhelming performance on high IOPS PCIe
> > SSDs. We know of some bottlenecks and ideas to test, but hoping to work
> > out the kinks in those ideas before Vault in two weeks.
> 
> Related to merging, or?
> 
> --
> Jens Axboe