Arch Linux

TJH

Member

Registered: 2005-02-02

Posts: 3

NFS problems/weirdness, diskless arch

Hi there

Im trying to built a diskless arch box with root mounted over NFS. My problem is that the filesystem seems very unresponsive. All machines on my LAN is connected using 1GBit. Large reads and writes on the NFS share seems OK, about 50MB/s. But if I for example extract the kernel source tarball, it takes about 11 minutes.
Yesterday I discovered that the filesystem was very responsive while extracting the kernel source, so i tried to extract two kernel tarballs concurrently and they both completed in about 2 minutes??? Can somebody please explain what is is going on here?

My fstab looks like:
192.168.1.1:/diskless/tjh-wrk   / nfs     rsize=1024,wsize=1024    0 0

But 'nfsstat -m' reports:

/ from 192.168.1.1:/diskless/tjh-wrk
Flags: rw,vers=3,rsize=262144,wsize=262144,hard,nolock,proto=tcp,timeo=7,retrans=3,sec=sys,addr=192.168.1.1

Shouldn't my fstab options override the defaults?

I really hope that somebody can explain this weird behavior

Thanks

//Toke

TJH

Member

Registered: 2005-02-02

Posts: 3

Re: NFS problems/weirdness, diskless arch

Hi again

Now i have tried to do some benchmarking on the NFS share to make the problem more clear.
If I run the following ruby script in the background:

#!/usr/bin/ruby
while true
  `tar xvf linux-2.6.20.tar`
end

and benchmark using: 'bonnie++ -u root -d /root/' I get the following results(which seems OK):

Version  1.03       ------Sequential Output------ --Sequential Input- --Random-
                    -Per Chr- --Block-- -Rewrite- -Per Chr- --Block-- --Seeks--
Machine        Size K/sec %CP K/sec %CP K/sec %CP K/sec %CP K/sec %CP  /sec %CP
tjh-wrk          4G 22420  48 33297   6 12486   5 31232  86 52775   8 232.2   0
                    ------Sequential Create------ --------Random Create--------
                    -Create-- --Read--- -Delete-- -Create-- --Read--- -Delete--
              files  /sec %CP  /sec %CP  /sec %CP  /sec %CP  /sec %CP  /sec %CP
                 16  1435   4  1771   3  1505   4  1766   4  3107   3  1600   3
tjh-wrk,4G,22420,48,33297,6,12486,5,31232,86,52775,8,232.2,0,16,1435,4,1771,3,1505,4,1766,4,3107,3,1600,3

If I then stop the ruby script and run the benchmark again I get the following disappointing results:

Version  1.03       ------Sequential Output------ --Sequential Input- --Random-
                    -Per Chr- --Block-- -Rewrite- -Per Chr- --Block-- --Seeks--
Machine        Size K/sec %CP K/sec %CP K/sec %CP K/sec %CP K/sec %CP  /sec %CP
tjh-wrk          4G 22705  48 29159   5 12231   5 33314  91 50284   7 340.7   0
                    ------Sequential Create------ --------Random Create--------
                    -Create-- --Read--- -Delete-- -Create-- --Read--- -Delete--
              files  /sec %CP  /sec %CP  /sec %CP  /sec %CP  /sec %CP  /sec %CP
                 16   180   0   133   0   193   0   174   0   123   0   201   0
tjh-wrk,4G,22705,48,29159,5,12231,5,33314,91,50284,7,340.7,0,16,180,0,133,0,193,0,174,0,123,0,201,0

I really hope somebody can help me with this one..

//Toke