Re: [PATCH] nfsstat: reorder nfs4 stats for 2.6.38 and up

[Benny Halevy <bhalevy@panasas.com>]

On 2011-04-27 23:29, Trond Myklebust wrote:
> On Wed, 2011-04-27 at 23:24 +0300, Benny Halevy wrote:
>> On 2011-04-27 21:58, Trond Myklebust wrote:
>>> On Wed, 2011-04-27 at 21:52 +0300, Benny Halevy wrote:
>>>> On 2011-04-27 17:16, Trond Myklebust wrote:
>>>>> On Wed, 2011-04-27 at 07:50 +0300, Benny Halevy wrote:
>>>>>> On 2011-04-25 17:11, Chuck Lever wrote:
>>>>>>> Hey all-
>>>>>>>
>>>>>>> So what are we going to do when adding NFSv4.2 to this mix?  Will we
>>>>>>> then have to freeze both the NFSv4.1 and NFSv4.0 procedure API in the
>>>>>>> kernel?  Seems painful.
>>>>>>
>>>>>> Good question.
>>>>>> How about changing the stat pseudo-file format to include an op
>>>>>> identifier along with its respective counter, printing a line per op?
>>>>>
>>>>> We already have that in /proc/self/mountstats
>>>>>
>>>>>
>>>>
>>>> You mean /proc/self/status?
>>>
>>> No.
>>
>> So can you please explain what you meant by the /proc/self/mountstats
>> example?
>>
>> This is what I see:
>>
>> $ head -3 /proc/self/mountstats
>> device rootfs mounted on / with fstype rootfs
>> device /proc mounted on /proc with fstype proc
>> device /sys mounted on /sys with fstype sysfs
>>
>> Benny
> 
> Try mounting an NFS partition. When I do, I also get:
> 
> device ila.local:/raid0/data/vol0 mounted on /net/ila.local/raid0/data/vol0 with fstype nfs statvers=1.0
> 	opts:	rw,vers=3,rsize=32768,wsize=32768,namlen=255,acregmin=3,acregmax=60,acdirmin=30,acdirmax=60,hard,proto=tcp,timeo=600,retrans=2,sec=sys,mountaddr=192.168.1.20,mountvers=3,mountport=993,mountproto=udp
> 	age:	3
> 	caps:	caps=0x3fcf,wtmult=4096,dtsize=4096,bsize=0,namlen=255
> 	sec:	flavor=1,pseudoflavor=1
> 	events:	4 101 0 0 6 0 107 0 0 0 0 0 12 0 0 0 0 0 0 0 0 0 0 0 0 
> 	bytes:	0 0 0 0 0 0 0 0 
> 	RPC iostats version: 1.0  p/v: 100003/3 (nfs)
> 	xprt:	tcp 826 1 1 0 2 24 24 0 24 0
> 	per-op statistics
> 	        NULL: 0 0 0 0 0 0 0 0
> 	     GETATTR: 6 6 0 752 672 0 1 1
> 	     SETATTR: 0 0 0 0 0 0 0 0
> 	      LOOKUP: 0 0 0 0 0 0 0 0
> 	      ACCESS: 5 5 0 692 600 0 1 1
> 	    READLINK: 0 0 0 0 0 0 0 0
> 	        READ: 0 0 0 0 0 0 0 0
> 	       WRITE: 0 0 0 0 0 0 0 0
> 	      CREATE: 0 0 0 0 0 0 0 0
> 	       MKDIR: 0 0 0 0 0 0 0 0
> 	     SYMLINK: 0 0 0 0 0 0 0 0
> 	       MKNOD: 0 0 0 0 0 0 0 0
> 	      REMOVE: 0 0 0 0 0 0 0 0
> 	       RMDIR: 0 0 0 0 0 0 0 0
> 	      RENAME: 0 0 0 0 0 0 0 0
> 	        LINK: 0 0 0 0 0 0 0 0
> 	     READDIR: 0 0 0 0 0 0 0 0
> 	 READDIRPLUS: 7 7 0 1112 13832 0 35 36
> 	      FSSTAT: 1 1 0 104 84 0 1 1
> 	      FSINFO: 2 2 0 208 160 0 0 0
> 	    PATHCONF: 1 1 0 104 56 0 0 0
> 	      COMMIT: 0 0 0 0 0 0 0 0
> 
> Providing this kind of statistic was _exactly_ the reason for adding
> mountstats in the first place.

Ah, cool!
This makes much more sense now :)

Benny