Re: Severe netbsd-6 NFS server-side performance issues

[Hauke Fath <hauke%Espresso.Rhein-Neckar.DE@localhost>]

At 22:55 Uhr -0700 30.5.2012, Brian Buhrow wrote:
>Another thought I had while looking at the bug report for
>Kern/46136.  What if you swap to something other than the raid controller?

The machine has 4 GB of RAM, and while it runs Netatalk and Samba besides
nfsd, they're not taxed much. top(1) says

Memory: 2298M Act, 1131M Inact, 14M Wired, 25M Exec, 3384M File, 23M Free
Swap: 1024M Total, 1024M Free

>       Also, I'm not sure about this, but it looks like the driver is
>multithreaded with the older raid controllers, but with the newer ones,
>it's not.  I'm not sure if the one you have qualifies as old or new from
>the driver's perspective.

The controller bios is dated 2006 - I guess that qualifies as 'old'?  ;)

>Christoz's comments about async disks also seems worth looking at.  Any
>chance to look at the mode pages for each disk?

Not from the controller bios - the configuration interface is one of the
reasons (besides the support punting my questions) why I wouldn't buy LSI
MegaRAIDs again. You can set the interface speed per disk, but IIRC it
doesn't tell you what the disk actually negociates for.

But, as I said, you probably wouldn't see the bonnie performance from disks
running async.

        hauke


--
"It's never straight up and down"     (DEVO)