mpls-linux-general

Re: [mpls-linux-general] Problem with CPU overload in older PCs

[lucapilosu\@libero\.it <luc...@li...>]

Yes, it is disabled!


---------- Initial Header -----------

>From      : "Adrian Popa" adr...@gm...
To          : "luc...@li..." luc...@li...
Cc          : "mpls-linux-general" mpl...@li...
Date      : Mon, 19 May 2008 13:16:25 +0300
Subject : Re: [mpls-linux-general] Problem with CPU overload in older PCs







> Did you disable debugging in mpls? Something like echo '0' >
> /sys/mpls/debug. Once you do that performance increases very much!
> 
> 
> On Mon, May 19, 2008 at 1:09 PM, luc...@li... <luc...@li...>
> wrote:
> 
> > Hello,
> > I'm doing some measures with traffic generators within a MPLS network.
> > My first LER is a Pentium 3 @647 MHz, and if I run the command "top" on it,
> > I see that the CPU load grows rapidly ove 96% as soon as I start sending
> > traffic (at the moment I'm sending 8Mbps, due to ksoftirqd (what is it?)
> > resulting in dropping packets from my flow without a predictable principle.
> > Do you know if it could be MPLS that boosts the CPU load this way?
> > Thanks,
> > Luca
> >
> >
> > -------------------------------------------------------------------------
> > This SF.net email is sponsored by: Microsoft
> > Defy all challenges. Microsoft(R) Visual Studio 2008.
> > http://clk.atdmt.com/MRT/go/vse0120000070mrt/direct/01/
> > _______________________________________________
> > mpls-linux-general mailing list
> > mpl...@li...
> > https://lists.sourceforge.net/lists/listinfo/mpls-linux-general
> >
> 

Re: [mpls-linux-general] Problem with CPU overload in older PCs

[lucapilosu\@libero\.it <luc...@li...>]

Solved!
My problem was caused by 2 WiFi channels that were too close to each other...(don't ask me why two 802.11a channels are overlapping!!).
Thanks a lot all the same.
Luca

Re: [mpls-linux-general] Problem with CPU overload in older PCs

[James R. L. <jl...@mi...>]

I'm glad you worked out your issue.  But just so you know there is a
performance bottle neck when you try to using interfaces in the 1Gbs and
10 Gbs range.  I'm working with another user to try and identify the
issues.  If you move your testing to higher speed interfaces let me
know and I can include you in the research.

On Mon, May 19, 2008 at 03:26:05PM +0200, luc...@li... wrote:
> Solved!
> My problem was caused by 2 WiFi channels that were too close to each other...(don't ask me why two 802.11a channels are overlapping!!).
> Thanks a lot all the same.
> Luca
> 
> 
> -------------------------------------------------------------------------
> This SF.net email is sponsored by: Microsoft 
> Defy all challenges. Microsoft(R) Visual Studio 2008. 
> http://clk.atdmt.com/MRT/go/vse0120000070mrt/direct/01/
> _______________________________________________
> mpls-linux-general mailing list
> mpl...@li...
> https://lists.sourceforge.net/lists/listinfo/mpls-linux-general

-- 
James R. Leu
jl...@mi...

Re: [mpls-linux-general] Problem with CPU overload in older PCs

[Chris R. <Chr...@nr...>]

James

Speaking of..... did you see my last request for the development tree?   
Sorry for being pushy on this... idle tester here.....

As an update, the 10GE configurations are in place and hope to have the 
20G Infinband systems in place by weeks end.  Also, we are building a 
slightly modified 10GE configuration from the baseline with PEs 
interfacing to CEs over multiple interfaces which are then either bond 
into a single aggregate flow or configured with traffic control 
signaling to deliver various QoS managed flows.  More things we need to 
see how the MPLS code responds too.

....chris

James R. Leu wrote:
> I'm glad you worked out your issue.  But just so you know there is a
> performance bottle neck when you try to using interfaces in the 1Gbs and
> 10 Gbs range.  I'm working with another user to try and identify the
> issues.  If you move your testing to higher speed interfaces let me
> know and I can include you in the research.
>
> On Mon, May 19, 2008 at 03:26:05PM +0200, luc...@li... wrote:
>   
>> Solved!
>> My problem was caused by 2 WiFi channels that were too close to each other...(don't ask me why two 802.11a channels are overlapping!!).
>> Thanks a lot all the same.
>> Luca
>>
>>
>> -------------------------------------------------------------------------
>> This SF.net email is sponsored by: Microsoft 
>> Defy all challenges. Microsoft(R) Visual Studio 2008. 
>> http://clk.atdmt.com/MRT/go/vse0120000070mrt/direct/01/
>> _______________________________________________
>> mpls-linux-general mailing list
>> mpl...@li...
>> https://lists.sourceforge.net/lists/listinfo/mpls-linux-general
>>     
>
>   
> ------------------------------------------------------------------------
>
> -------------------------------------------------------------------------
> This SF.net email is sponsored by: Microsoft 
> Defy all challenges. Microsoft(R) Visual Studio 2008. 
> http://clk.atdmt.com/MRT/go/vse0120000070mrt/direct/01/
> ------------------------------------------------------------------------
>
> _______________________________________________
> mpls-linux-general mailing list
> mpl...@li...
> https://lists.sourceforge.net/lists/listinfo/mpls-linux-general
>   

-- 
Christopher Robson
Senior Computer Scientist, GS-15
Naval Research Laboratory
Center for Computational Science
Networking, Code 5591
4555 Overlook ave.
Washington, D.C. 20375-5320
(COM)  202-404-3138
(VoIP) 2024043138@ATDNet
(CHAT) Chris.Robson@ATDNet

Re: [mpls-linux-general] Problem with CPU overload in older PCs

[Adrian P. <adr...@gm...>]

[offtopic]
Aren't 802.11 channels supposed to overlap?

Quote from wikipedia:
http://en.wikipedia.org/wiki/IEEE_802.11#Channels_and_international_compatibility

802.11 divides each of the above-described bands into channels, analogously
to how radio and TV broadcast bands are carved up but with greater channel
width and overlap. For example the 2.4000-2.4835 GHz band is divided into 13
channels each of width 22 MHz but spaced only 5 MHz apart, with channel 1
centered on 2412 MHz and 13 on 2472, to which Japan adds a 14th channel 12
MHz above channel 13.

True, this only applies to 2.4GHz bandwidth, not to 5GHz used by 802.11a...
[/offtopic]

On Mon, May 19, 2008 at 4:26 PM, luc...@li... <luc...@li...>
wrote:

> Solved!
> My problem was caused by 2 WiFi channels that were too close to each
> other...(don't ask me why two 802.11a channels are overlapping!!).
> Thanks a lot all the same.
> Luca
>
>

Re: [mpls-linux-general] Problem with CPU overload in older PCs

[lucapilosu\@libero\.it <luc...@li...>]

I would be happy to give some help, but I'm working with wireless links, so I will not reach that speeds!!



---------- Initial Header -----------

>From      : "James R. Leu" jl...@mi...
To          : "luc...@li..." luc...@li...
Cc          : "adrian.popa.gh" adr...@gm...,"mpls-linux-general" mpl...@li...
Date      : Mon, 19 May 2008 08:29:50 -0500
Subject : Re: [mpls-linux-general] Problem with CPU overload in older PCs







> I'm glad you worked out your issue.  But just so you know there is a
> performance bottle neck when you try to using interfaces in the 1Gbs and
> 10 Gbs range.  I'm working with another user to try and identify the
> issues.  If you move your testing to higher speed interfaces let me
> know and I can include you in the research.
> 
> On Mon, May 19, 2008 at 03:26:05PM +0200, luc...@li... wrote:
> > Solved!
> > My problem was caused by 2 WiFi channels that were too close to each other...(don't ask me why two 802.11a channels are overlapping!!).
> > Thanks a lot all the same.
> > Luca
> > 
> > 
> > -------------------------------------------------------------------------
> > This SF.net email is sponsored by: Microsoft 
> > Defy all challenges. Microsoft(R) Visual Studio 2008. 
> > http://clk.atdmt.com/MRT/go/vse0120000070mrt/direct/01/
> > _______________________________________________
> > mpls-linux-general mailing list
> > mpl...@li...
> > https://lists.sourceforge.net/lists/listinfo/mpls-linux-general
> 
> -- 
> James R. Leu
> jl...@mi...
> 

Re: [mpls-linux-general] Problem with CPU overload in older PCs

[lucapilosu\@libero\.it <luc...@li...>]

802.11a channels (in 5 GHz band) are supposed to be non-overlapping, but with a spectrum analyzer I saw that my board's channels spectrum is not very sharp, but they have some tails that partially overlap with the "neighbours".


---------- Initial Header -----------

>From      : "Adrian Popa" adr...@gm...
To          : "luc...@li..." luc...@li...
Cc          : "mpls-linux-general" mpl...@li...
Date      : Tue, 20 May 2008 18:35:02 +0300
Subject : Re: [mpls-linux-general] Problem with CPU overload in older PCs







> [offtopic]
> Aren't 802.11 channels supposed to overlap?
> 
> Quote from wikipedia:
> http://en.wikipedia.org/wiki/IEEE_802.11#Channels_and_international_compatibility
> 
> 802.11 divides each of the above-described bands into channels, analogously
> to how radio and TV broadcast bands are carved up but with greater channel
> width and overlap. For example the 2.4000-2.4835 GHz band is divided into 13
> channels each of width 22 MHz but spaced only 5 MHz apart, with channel 1
> centered on 2412 MHz and 13 on 2472, to which Japan adds a 14th channel 12
> MHz above channel 13.
> 
> True, this only applies to 2.4GHz bandwidth, not to 5GHz used by 802.11a...
> [/offtopic]
> 
> On Mon, May 19, 2008 at 4:26 PM, luc...@li... <luc...@li...>
> wrote:
> 
> > Solved!
> > My problem was caused by 2 WiFi channels that were too close to each
> > other...(don't ask me why two 802.11a channels are overlapping!!).
> > Thanks a lot all the same.
> > Luca
> >
> >
>