Re: **SPAM: Re: [Madwifi-users] Changing bin interval and txpower: r1454

[Tom S. <tsh...@qo...>]

You'd need to run that test for considerably longer than 180 seconds. I'd 
suggest running it overnight. We have seen some cases of hard freezes when 
running continuous data (e.g. streaming video) through a pair of associated 
ad-hoc nodes over, say, a 24 hour period. Dmesg has been unrevealing. We use 
a hardware watchdog to deal with the freeze but of course a true fix would 
be better.

Tom Sharples
President
Qorvus Systems, Inc.
www.qorvus.net
360.243.7371


----- Original Message ----- 
From: "Vikram KAUL" <vk...@re...>
To: "John Clark" <jc...@me...>
Cc: <mad...@li...>
Sent: Thursday, March 02, 2006 9:54 AM
Subject: **SPAM: Re: [Madwifi-users] Changing bin interval and txpower: 
r1454


>
> Thanks for the reply, please see my comments
>
> jclark>A couple of items. I have been using a very early version of 
> Madwifi
> jclark>which supported ad-hoc mode. I have not been
> jclark>able to get the same reliability out of anything 'recent', with 
> either
> jclark>the NG or the 'old' driver. I have with the May 2005
> jclark>version, which was pretty 'early', and probably has more problems 
> than I
> jclark>could imagine, but it works reliably in ad-hoc
> jclark>mode which is the name of my game.
>
> Is there some place where the problems are specifically documented ? I
> have been experimenting with it locally and have had no such problems,
> although the conditions where those occur might not happened with me
>
> jclark>
> jclark>The last version of Madwifi-ng that I checked was about 2 weeks 
> ago, same
> jclark>set of symptoms, which was as soon as
> jclark>any really net traffic begins, the driver runs through malloc'd 
> memory,
> jclark>and the kernel issues diagnostics. Sometimes
> jclark>taking the interface 'down', via 'ifconfig ath0 down', then back up
> jclark>'cures' the problem, sometimes it just panics and stops.
> jclark>
>
> Is there a specific amount/rate of traffic that does this ? I was
> stress-testing the link (only for throughput measurements) and did not
> encounter the problems you mention.
>
> Test: netperf for UDP_STREAM for 1400 bytes for 180 seconds gave me about
> 5.2 Mbps throughput and no panic situations. The nodes were close to each
> other (at about 50~55/94 quality as reported by iwspy)
>
> I am using cisco-aironet cards and the r1454 version of the driver.
>
> I even tried to run the same test with UDP_STREAM test from and to
> the same node from two differen nodes. (All three ad-hoc of
> course). The throughput did drop, but still I did not get the panic
> conditions.
>
> Can someone point specifically to problems that exist with ad-hoc mode in
> madwifi ?
>
>
> Thanks
> Vikram
>
>
> -------------------------------------------------------
> This SF.Net email is sponsored by xPML, a groundbreaking scripting 
> language
> that extends applications into web and mobile media. Attend the live 
> webcast
> and join the prime developer group breaking into this new coding 
> territory!
> http://sel.as-us.falkag.net/sel?cmd=lnk&kid=110944&bid=241720&dat=121642
> _______________________________________________
> Madwifi-users mailing list
> Mad...@li...
> https://lists.sourceforge.net/lists/listinfo/madwifi-users