Re: [Linux-cachefs] Cachefs fails to take effect on nfs4?

[Rince <rincebrain gmail com>]

Fascinating.

I sit corrected. This only occurs for recursively submounted
directories, e.g. /scratch/subdir1.
cat /proc/fs/fscache/stats > file1
pv < /scratch/subdir1/testfile > /dev/null
[50 MB of data fly by]
cat /proc/fs/fscache/stats > file2

diff file1 file2
[no output]

mount | grep /scratch
manticore:/scratch on /scratch type nfs4
(rw,fsc,clientaddr=128.220.X.X,addr=128.220.X.Y)


- Rich

On Thu, May 21, 2009 at 10:09 AM, David Howells <dhowells redhat com> wrote:
> Rince <rincebrain gmail com> wrote:
>
>> If I mount /scratch with -t nfs4 -o rw,fsc, and do ~50 MB of IO, I get the
>> same:
>
> Can you describe the IO you're doing?  Reading 1 x 50MB file?  Writing 1000000
> 50B files?
>
>> So...what's going on here?  Am I missing something, or is -t nfs4
>> somehow making it magically ignored?
>
> NFS4 should work with the same effectiveness as NFS2 or 3.
>
> David
>
> --
> Linux-cachefs mailing list
> Linux-cachefs redhat com
> https://www.redhat.com/mailman/listinfo/linux-cachefs
>



-- 

New systems generate new problems.