Hey Rahul,
I do not have bandwidth for doing the test your asked,
You haven't answered the questions asked, so please answer
them first before we can have this conversation continuing,
1. Did you use vm or bare metal machine for testing,
2. Did you make any modification to ovs, if so, what did y
Hello Alex,
Thanks for your immediate reply.
Alex,
Can you please test the same performance with the same flow by enabling 2
cores of the server only.And share the performance report with me of OVS VS
Kernel Bridge.
On Wed, Dec 10, 2014 at 1:10 PM, Alex Wang wrote:
> Hey Rahul,
>
> The war
Hey Rahul,
The warning means the handler threads was busying running for 16 seconds
before calling time_poll() again (in other words, before trying to sleep
again).
The log showed suspiciously long system running. Definitely not expected,
Could you confirm the ovs kernel module version via 'mod
These logs are coming from *lib/timeval.c *under the function
*log_poll_interval* in OVS source code.
On Wed, Dec 10, 2014 at 12:17 PM, Rahul Arora
wrote:
> Alex,
>
> I am getting these logs when i am getting 100% CPU usage.
>
> Please have a look on the RED part.
>
> ==
Alex,
I am getting these logs when i am getting 100% CPU usage.
Please have a look on the RED part.
==
2014-11-25T06:04:37.604Z|00095|timeval(handler5)|WARN|Unreasonably long
16127ms poll interval (260ms user, 17320ms system)
2014-11-25T06:04:37.604Z|00096|tim
I believe it has 16 cores, with hyperthreading disabled,
On Tue, Dec 9, 2014 at 9:37 PM, Rahul Arora
wrote:
> Hi Alex,
>
> The platform which you are using for your performance testing is of how
> many cores??
>
> On Tue, Dec 9, 2014 at 10:45 PM, Alex Wang wrote:
>
>> Hey Rahul,
>>
>> The kerne
Hi Alex,
The platform which you are using for your performance testing is of how
many cores??
On Tue, Dec 9, 2014 at 10:45 PM, Alex Wang wrote:
> Hey Rahul,
>
> The kernel version should not make a difference,
>
> I'm curious, since ovs-2.3.0 is multithreaded, not sure how do you measure
> the
Hey Rahul,
The kernel version should not make a difference,
I'm curious, since ovs-2.3.0 is multithreaded, not sure how do you measure
the cpu usage to be 100%?
Also, could you post the full `mpstat -P ALL` output, so we can see the
distribution of cpu usage,
Thanks,
Alex Wang,
On Tue, Dec
Hi Alex,
Thanks for your help.
I am using the following command to test CPU usage.
mpstat -P ALL 1
I was using 3.12 kernel version and you are using 3.13.0-30-generic
version.Please let me know if this can be the issue??
On Sat, Dec 6, 2014 at 5:04 AM, Alex Wang wrote:
> Hey Rahul,
>
> Cou
Hey Rahul,
Could you help confirm the following:
1. did you use vm or bare metal machine for testing?
2. how do you measure the cpu utilization of ovs and kernel bridge?
We ran some throughput tests, and we observed constant better
performance over the Linux bridge.
Below is a set of our own
I'm not sure, but may be using a multi-cpu-supporting netowrk card can
help. If you using something like e1000 or bmx, your network performance
is limited by single thread. Modern network cards (like ixgb/ixgbe) can
utilize more cpu.
P.S. In case of small packets it's better to count pps, not
Hi Team,
We are doing comparison of throughput and CPU consumption between OVS 2.3.0
and kernel bridge with different packet size.
We are observing huge difference in performance. With frame size 64 and 128
bytes unidirectional traffic from port1 to port2 below are the numbers.
*OVS 2.3.0 (Dual
12 matches
Mail list logo
