[Gluster-users] This benchmark is OK?

[soft_lawrency at hotmail.com]

Hi Amar,
     here is my benchmark, pls help me to evaluate it.
     1. [Env - Storage servers]
     Gluster version: 3.3 beta3
     OS : CentOS 6.1
     2* Server : CPU :         E5506  @ 2.13GHz 4*core
                       MEM:       16G
                       Disk :        1T * 12
                       Net:          bond0 = 4 * 1Gb
     2. [Env - Testing clients]
     OS: CentOS 6.1
     6* Server: CPU:           5150  @ 2.66GHz 
                     MEM:          8G
                     Disk:            RAID0 = 1T * 3 
                     Net:             bond0 = 2 * 1Gb
     3. [Env - Switch]
     1 * H3C Switch
     4. [Env - Testing Tools]
     iozone for cluster, clients is rsh cluster.
     5. Volume info 
     Type: distributed - stripe
     bricks: 6 * 4 = 24
     6. Iozone command: 
     ./iozone -r 1m -s 50g -t 25 -i 0 -i 1 -+m /CZFS/client_list -R -b report.xls -c -C -+k -e 

      then my benchmark is :
     
     "  Initial write "  424258.41 KB

     "        Rewrite "  458250.97 KB

     "           Read "  671079.30 KB

     "        Re-read "  797509.20 KB
 
     here I have 4 Questions:
     Q1: Why the Read is better than write, is ok?
     Q2: Refer to experience, This benchmark is best , good or bad? 
     Q3: How to optimize the gluster to impove the benchmark?
     Q4: How can I get other friends' benchmark, which I use to compared?
     
      Tks very much.
      
      Regards,
      Lawrence.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://supercolony.gluster.org/pipermail/gluster-users/attachments/20120517/970465b2/attachment.html>