Re: RAID stripe size

To: tech-kern%netbsd.org@localhost
Subject: Re: RAID stripe size
From: Edgar Fuß <ef%math.uni-bonn.de@localhost>
Date: Mon, 6 Jun 2011 16:02:15 +0200

> The filesystem block size (or, where this is not possible, the maximum
> cluster size the filesystem will write) should be equal to sectPerSU
> times the number of data (not parity) disks. 
Yes. That's what I would call ``stripe size''. Is that the wrong term?
I have sectPerSU=32, so with a three-disc RAID 5, that's 32*512*(3-1)=32k per 
what-I-would-call-stripe, and that's what I chose as fbsize. Did I mess up 
something?

References:
- Re: 5.1 RAID5 write performance
  - From: Michael L. Hitch
- Re: 5.1 RAID5 write performance
  - From: Thor Lancelot Simon
- RAID stripe size (was: 5.1 RAID5 write performance)
  - From: Edgar Fuß
- Re: RAID stripe size (was: 5.1 RAID5 write performance)
  - From: Thor Lancelot Simon

Prev by Date: Re: RAID stripe size (was: 5.1 RAID5 write performance)
Next by Date: Re: RAID stripe size (was: 5.1 RAID5 write performance)
Previous by Thread: Re: RAID stripe size (was: 5.1 RAID5 write performance)
Next by Thread: Re: RAID stripe size (was: 5.1 RAID5 write performance)
Indexes:

Home | Main Index | Thread Index | Old Index