[PATCH RFC 0/2] nvme-multipath: additional iopolicies

[Hannes Reinecke]

On 7/26/23 15:25, Christoph Hellwig wrote:
> On Wed, Jul 26, 2023 at 03:23:03PM +0200, Hannes Reinecke wrote:
>> Hi all,
>>
>> as Ewan raised the issue that we are suffering from a buffer-bloat
>> problem when using the round-robin I/O scheduler here are additiona
>> I/O schedulers to improve the situation.
>> The schedulers are untested, and are compile-tested only.
>> I just putting them out here as I had several requests for them;
>> I'm happily improving them upon feedback.
> 
> FYI, the policy for nvme-multpath has mostly been that we want to
> see very solid used cases.  So in addition to actually testing
> the code, I'd like to see detailed and solid use cases.  And preferably
> just one new policy and not three if possible..
> 
I could not agree more.
That was precisely why I was reluctant to post them in the first place.
But I have been asked by several parties and really don't have time now 
to do a proper testing (what with me being in the middle of a server 
room move).
So I decided to put them out as a RFC, fully expecting things to change.
And also expecting to have some solid data from testing to check which 
I/O scheduler really improves the situation.

Cheers,

Hannes
-- 
Dr. Hannes Reinecke                Kernel Storage Architect
hare at suse.de                              +49 911 74053 688
SUSE Software Solutions GmbH, Maxfeldstr. 5, 90409 Nürnberg
HRB 36809 (AG Nürnberg), Geschäftsführer: Ivo Totev, Andrew
Myers, Andrew McDonald, Martje Boudien Moerman