Re: Disk I/O slowness

Stefan Monnier <[email protected]>
Newsgroups gmane.linux.debian.user
Message-ID <[email protected]>
Michael Stone [2026-08-16 18:12:32] wrote:
> On Wed, Aug 05, 2026 at 09:44:42PM -0400, Stefan Monnier wrote:
>>>> RAID-5 and RAID-6 rebuild of 8 drive array would take a very long time,
>>>> because not only it will have to recover (not just copy) the data from
>>>> error correction data spread across all drives
>>> Reading the data from all the drives is a positive rather than a negative:
>>> it means the controller is reading from say, 7 drives to build data on
>>> 1.  That means a lot more idle cycles on the 7 as opposed to having 1 drive
>>> 100% busy (which matters if you're trying to use the array during the
>>> rebuild).
>>But you need to read all "other blocks" to compute a given missing
>>block, so it takes 6x more reading to rebuild your array than with
>>a raid 1/10.
> Who cares; if it takes less time to read those blocks, why does it matter?

But does it take less time?  AFAIK it takes more time:
it might take about the same *wallclock* time because all the disks can be
read concurrently, but only if there's no other activity and only if the
rest of the hardware can keep up.

> The limitations on modern hardware aren't the same as they were 30
> years ago.

But the basics remain the same: with RAID 1/10 rebuilding a disk
requires reading 1 full disk and writing it to the new disk, whereas
with RAID 5, it requires reading all the data on all the remaining disks
and then combining the result and writing it to the new disk, so I can't
see how that can mean "a lot more idle cycles".


=== Stefan
lmpx.com only provides a reader for public news (NNTP) servers. It is not affiliated with the servers or forums shown here and is not responsible for the content of articles, which is written by their respective authors.