>> RAIDframe operates on units called stripes. [...RMW...]
> Your description matches a paragraph in raidctl(8) that starts with
> 'Tuning RAID 5 sets is trickier.' while my setup is RAID 1
Ah, that's important information.
> and recommended values for SectPerSU are 32 to 128. I set it to 128
Edgar Fu? wrote:
> RAIDframe operates on units called stripes. FFS mostly operates on FS blocks.
> If one FS block is not exactly a whole number of stripes (either because of
> misalignment or because an FS block is smaller than a RAID stripe), then each
> write of an FS block will force the
> Can you please elaborate? Do they have to match for the best performance,
> or should one be a multiple of the other?
RAIDframe operates on units called stripes. FFS mostly operates on FS blocks.
If one FS block is not exactly a whole number of stripes (either because of
misalignment or
On 11/17/17 07:44, Ryota Ozaki wrote:
On Wed, Nov 15, 2017 at 1:53 PM, Ryota Ozaki wrote:
On Fri, Nov 10, 2017 at 6:35 PM, Ryota Ozaki wrote:
Hi,
http://www.netbsd.org/~ozaki-r/IFEF_MPSAFE.diff
I'm going to commit the above change that
Edgar Fu? wrote:
> > when raid is configured for the first time, it needs to do something
> > with every block?
> raidctl -i, yes.
That's what I thought. Thanks for confirming.
> Other than that, you'll need to align your FS blocks with RAID stripes
> and each component's stripe part with
> when raid is configured for the first time, it needs to do something
> with every block?
raidctl -i, yes.
Other than that, you'll need to align your FS blocks with RAID stripes
and each component's stripe part with whatever cgd operates on.
Why do you do RAID-on-CGD and not CGD-on-RAID?
Mouse wrote:
> > My raid is made of two disks dk8 and dk9. When there is an activity
> > on cgd, dk8 and dk9 shows more than 2x bytes each than cgd. Why is
> > that?
>
> Speculation: read-modify-write cycles?
>
> Speculation: cgd's encryption leading to accessing more of its
> underlying disk
> My raid is made of two disks dk8 and dk9. When there is an activity
> on cgd, dk8 and dk9 shows more than 2x bytes each than cgd. Why is
> that?
Speculation: read-modify-write cycles?
Speculation: cgd's encryption leading to accessing more of its
underlying disk than the upper-layer access?
Alexander Nasonov wrote:
> My raid is made of two disks dk8 and dk9. When there is an activity on cgd,
> dk8 and dk9 shows more than 2x bytes each than cgd. Why is that?
>
>wd01525 71M 66.8 msix2 vec 0 2730
> ftarg
>wd11524 71M 59.3
My raid is made of two disks dk8 and dk9. When there is an activity on cgd,
dk8 and dk9 shows more than 2x bytes each than cgd. Why is that?
$ systat 2
:vmstat
Disks: seeks xfers bytes %busy msix1 vec 4 2048 fmin
wd01525 71M 66.8 msix2
10 matches
Mail list logo