Bug 882088 - v7's network should reflect the max real transfer rate/performace
Summary: v7's network should reflect the max real transfer rate/performace
Keywords:
Status: CLOSED DUPLICATE of bug 882093
Alias: None
Product: Red Hat Hardware Certification Program
Classification: Retired
Component: Test Suite (tests)
Version: 6.1
Hardware: Unspecified
OS: Unspecified
unspecified
unspecified
Target Milestone: ---
: ---
Assignee: Greg Nichols
QA Contact: Red Hat Kernel QE team
URL:
Whiteboard:
Depends On:
Blocks:
TreeView+ depends on / blocked
 
Reported: 2012-11-30 05:01 UTC by QinXie
Modified: 2012-12-19 16:50 UTC (History)
4 users (show)

Fixed In Version:
Doc Type: Bug Fix
Doc Text:
Clone Of:
Environment:
Last Closed: 2012-12-19 16:50:41 UTC
Embargoed:


Attachments (Terms of Use)

Description QinXie 2012-11-30 05:01:30 UTC
In review process, reviewer need to check bw_tcp value to see the real transfer rate. If the transfer rate is much lower than the naming speed, we need to double check with vendor.

In cert#828100, v7 report the transfer rate for QDR device is about 4MBps:
TCP latency using 10.0.0.10: 0.0940 microseconds
testing bandwidth to 10.0.0.10...
0.065536 408.99 MB/sec


And vendor reports when he run tests manually as following steps, return value of bw_tcp can reach 2.1GBps. And iperf return value is 27 MBit/sec. 
[root@SA2260-X9DR3-LN4F ~]# bw_tcp -m 65520 -P 8 10.0.0.10
0.065520 2166.37 MB/sec
For detail please refer to https://hardware.redhat.com/show.cgi?id=828100#c45

steps is belowing, for detail please refer to https://hardware.redhat.com/show.cgi?id=828100#c47:
- set CONNECTED_MODE=yes and MTU=65520 in
/etc/sysconfig/network-scripts/ifcfg-ib0 (for both server and test system)
- set the governor to performance in /etc/sysconfig/cpuspeed
- bound the process qib_cq to a specific CPU core with 'taskset -pc 3 $(pidof
qib_cq)' on both server and client

Can network test be improved to reflect max real value?


Note You need to log in before you can comment on or make changes to this bug.