Re: NFS client (10x) performance regression 2.6.14.7 -> 2.6.15

[Trond Myklebust]

On Fri, 2006-03-31 at 14:45 +0200, Jakob Oestergaard wrote:
> On Fri, Mar 31, 2006 at 07:22:50AM -0500, Trond Myklebust wrote:
> ...
> > 
> > Some nfsstat output comparing the good and bad cases would help.
> 
> Clean boot on 2.6.15 and 2.6.14.7, one run of nfsbench with
> LEADING_EMPTY_SPACE=1.  I've skipped the NFS v2 stats because they're
> all 0.

Why all the GETATTR calls? Are you running with 'noac' set?

I don't have a 2.6.15 kernel to run with, but on a recent git pull, I
get a total of 6 GETATTR calls when I run your nfsbench program.

The number of READ calls is 1, and the number of WRITE calls is 161 (I'm
running with 64k wsize).

Cheers,
  Trond

> --- Run on bad kernel ---
> 
> [puffin:joe] $ uname -a
> Linux puffin 2.6.15 #1 SMP Fri Mar 31 11:10:28 CEST 2006 i686 GNU/Linux
> [puffin:joe] $ nfsstat 
> Client rpc stats:
> calls      retrans    authrefrsh
> 63         0          0       
> 
> <snip v2 stats>
> 
> Client nfs v3:
> null       getattr    setattr    lookup     access     readlink   
> 0       0% 11     18% 0       0% 26     42% 14     22% 0       0% 
> read       write      create     mkdir      symlink    mknod      
> 4       6% 0       0% 0       0% 0       0% 0       0% 0       0% 
> remove     rmdir      rename     link       readdir    readdirplus
> 0       0% 0       0% 0       0% 0       0% 0       0% 4       6% 
> fsstat     fsinfo     pathconf   commit     
> 0       0% 2       3% 0       0% 0       0% 
> [puffin:joe] $ time ./nfsbench 
> 
> real    0m29.242s
> user    0m0.000s
> sys     0m0.160s
> [puffin:joe] $ nfsstat 
> Client rpc stats:
> calls      retrans    authrefrsh
> 13240      0          0       
> 
> <snip v2 stats>
> 
> Client nfs v3:
> null       getattr    setattr    lookup     access     readlink   
> 0       0% 2583   19% 0       0% 30      0% 24      0% 0       0% 
> read       write      create     mkdir      symlink    mknod      
> 7045   53% 3200   24% 1       0% 0       0% 0       0% 0       0% 
> remove     rmdir      rename     link       readdir    readdirplus
> 0       0% 0       0% 0       0% 0       0% 0       0% 34      0% 
> fsstat     fsinfo     pathconf   commit     
> 0       0% 2       0% 0       0% 319     2% 
> 
> [puffin:joe] $
> 
> --- Run on good kernel ---
> 
> [puffin:joe] $ uname -a
> Linux puffin 2.6.14.7 #1 SMP Fri Mar 31 10:41:59 CEST 2006 i686
> GNU/Linux
> [puffin:joe] $ nfsstat 
> Client rpc stats:
> calls      retrans    authrefrsh
> 52         0          0       
> 
> <snip v2 stats>
> 
> Client nfs v3:
> null       getattr    setattr    lookup     access     readlink   
> 0       0% 7      14% 0       0% 24     48% 13     26% 0       0% 
> read       write      create     mkdir      symlink    mknod      
> 4       8% 0       0% 0       0% 0       0% 0       0% 0       0% 
> remove     rmdir      rename     link       readdir    readdirplus
> 0       0% 0       0% 0       0% 0       0% 0       0% 0       0% 
> fsstat     fsinfo     pathconf   commit     
> 0       0% 2       4% 0       0% 0       0% 
> 
> [puffin:joe] $ time ./nfsbench 
> 
> real    0m0.224s
> user    0m0.000s
> sys     0m0.072s
> [puffin:joe] $ nfsstat 
> Client rpc stats:
> calls      retrans    authrefrsh
> 384        0          0       
> 
> <snip v2 stats>
> 
> Client nfs v3:
> null       getattr    setattr    lookup     access     readlink   
> 0       0% 10      2% 1       0% 26      6% 15      3% 0       0% 
> read       write      create     mkdir      symlink    mknod      
> 6       1% 321    84% 0       0% 0       0% 0       0% 0       0% 
> remove     rmdir      rename     link       readdir    readdirplus
> 0       0% 0       0% 0       0% 0       0% 0       0% 0       0% 
> fsstat     fsinfo     pathconf   commit     
> 0       0% 2       0% 0       0% 1       0% 
> 
> [puffin:joe] $
> 
> 

-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/