Hey Max & Nathan,
On Wed, Oct 19, 2011 at 06:25:33AM +1100, Nathan Scott wrote:
> ----- Original Message -----
> > On Mon, 17 Oct 2011 13:38:38 -0500, Ben Myers wrote:
> >
> > bpm> Here is an nfs client pmda that reports statistics from
> > bpm> /proc/self/mountstats.
> >
> Max> Just shows that if you wait long enough someone will do your job for
> Max> you - I was thinking about adding these metrics to Linux PMDA but
> Max> got lazy doing string parsing in C. Perl is easier for working with
> Max> strings
Yep. I needed something quick so it ended up in perl. ;)
Max> - call details.
Max> It's easy to create a ueber-array of all options and just list
Max> them all but it's probably not the most convinient for the user. I was
Max> thinking about mapping some of the options so that different NFS
Max> versions can have similar names.
I'm not sure I understand. Can you give an example?
Maybe something like:
nfs.reqs.getattr.count
vs
nfs4.reqs.getattr.count
> Max> - U32 for counts is probably a bit low - any particular reason for
> Max> not going U64?
>
> Nathan> What type is it in the kernel? (need to match - there's also helper
> Nathan> pmda_long and pmda_ulong types if thats what the kernel is using)
Exactly. I used types to match those in the kernel. Hopefully I didn't
overlook any.
Max> - per-transport metrics - once again would be nice to have some
Max> commonality.
Hmmm. Again you lost me.
Max> - instance domain - any particular reason you've used server 'exports'
Max> and not local mount points?
Yeah, maybe you're right. My impression had been that separate nfs
client superblocks can share the same counters and transports if they're
using the same server... so that's why I had an instance per server
export.
Max> BTW, they're not really exports - one can
Max> mount any directory (or even a file) within exported hierarchy.
Sometimes I forget that too... force of habit. Whether a separate
instance is needed should depend upon whether the counters are shared by
the nfs mounts in that situation. That's something I need to look into.
Max> I'll give it a spin on a Linux box to see how it works.
THanks! It might be a little while before I can work on it again... but
I'll try to make use of the pointers you've provided.
Regards,
Ben
|