[prev in list] [next in list] [prev in thread] [next in thread]
List: linux-ide
Subject: Re: [PATCH 2 of 9] block: Export I/O topology for block devices and partitions
From: "Martin K. Petersen" <martin.petersen () oracle ! com>
Date: 2009-04-24 14:53:36
Message-ID: yq11vri9k7j.fsf () sermon ! lab ! mkp ! net
[Download RAW message or body]
>>>>> "Kay" == Kay Sievers <kay.sievers@vrfy.org> writes:
Kay> Wouldn't it be good to include "sector", like the queue files do?
Kay> The alignment of a partition could mean many things.
Kay> /sys/block/<disk>/sector_alignment
Kay> /sys/block/<disk>/<partition>/sector_alignment
Well, the whole point of this exercise is to get rid of the
about-to-become-incorrect notion of a "sector" :)
What I'd like to see is these values being picked up by libdisk and used
to replace the current hacks that extract MD stripe size, etc.
So the values exported are:
alignment - Use this to add padding.
minimum_io_size - Don't submit I/Os smaller than this. May be
bigger than both physical and logical block
sizes at the bottom of the stack.
optimal_io_size - Use this request size for best performance.
Think full stripe write.
Those values apply to all block device types and are adjusted when
stacking.
Kay> And prefixing the io values might be easier to read when they show
Kay> up in a group?
Kay> /sys/block/<disk>/queue/io_minimum_size
Kay> /sys/block/<disk>/queue/io_optimal_size
Kay> /sys/block/<disk>/queue/io_...
I don't really have immediate plans to add more.
I'm not married to any particular naming scheme but I felt that
minimum_io_size was more readable than io_minimum_size. Which I why I
kept the user-visible names as is.
--
Martin K. Petersen Oracle Linux Engineering
--
To unsubscribe from this list: send the line "unsubscribe linux-ide" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
[prev in list] [next in list] [prev in thread] [next in thread]
Configure |
About |
News |
Add a list |
Sponsored by KoreLogic