Skip site navigation (1)Skip section navigation (2)
Date:      Thu, 9 Jul 2009 11:36:51 +0900
From:      Weongyo Jeong <weongyo.jeong@gmail.com>
To:        Nikos Vassiliadis <nvass9573@gmx.com>
Cc:        freebsd-net@freebsd.org
Subject:   Re: ndis and USB wirelless ethernet
Message-ID:  <20090709023651.GA1553@weongyo.cdnetworks.kr>
In-Reply-To: <4A531A94.40701@gmx.com>
References:  <4A43386D.80500@gmx.com> <20090625103420.GD31161@weongyo.cdnetworks.kr> <4A436A8A.1000405@gmx.com> <20090626041246.GE31161@weongyo.cdnetworks.kr> <4A461AF9.7040900@gmx.com> <20090629032520.GA1138@weongyo.cdnetworks.kr> <4A4880EF.5010206@gmx.com> <4A4E2873.3010501@gmx.com> <20090706043747.GD1138@weongyo.cdnetworks.kr> <4A531A94.40701@gmx.com>

next in thread | previous in thread | raw e-mail | index | archive | help
On Tue, Jul 07, 2009 at 12:51:16PM +0300, Nikos Vassiliadis wrote:
> Weongyo Jeong wrote:
> >I'm happy to see your device is successfully associated with AP.
> >However it seems it's a bad news that you sometimes meet crashes.  Does
> >a random crash mean a OS hang (e.g. could not type any keys) or no more
> >work of network operations?
> 
> It hangs, I cannot use the keyboard and I have to power-cycle it.
> It can happen after some time downloading and uploading. It hangs
> after 5 to 30 minutes of heavy traffic. By heavy traffic, I mean
> the maximum I can get from this device, which is 50KBytes/sec.
> 
> I am not sure what will happen if I let it idle for, let's say
> one day, but I haven't had a single crash during times with
> low activity, such as ssh traffic.
> 
> >Frankly speaking, for both cases it looks I could not provide any
> >solutions without backtraces unless I encountered same problems on my
> >environment.  It'd better if we can reproduce its problem easily.
> 
> Unfortunately, I have no solid facts to show you. The only strange
> thing I've seen and is consistent, is this:
> 
> speed# vmstat -m | grep USBdev ; sleep 1 ; vmstat -m | grep USBdev ; 
> sleep 1 ; vmstat -m | grep USBdev
>        USBdev    53     4K       -   267579  16,32,128,1024
>        USBdev    53     4K       -   267612  16,32,128,1024
>        USBdev    53     4K       -   267642  16,32,128,1024
> speed#
> speed# vmstat -m | grep USBdev ; sleep 1 ; vmstat -m | grep USBdev ; 
> sleep 1 ; vmstat -m | grep USBdev
>        USBdev    53     4K       -   268071  16,32,128,1024
>        USBdev    53     4K       -   268101  16,32,128,1024
>        USBdev    53     4K       -   268140  16,32,128,1024
> 
> And then with some traffic:
> speed# ping -i 0.01 192.168.1.1 > /dev/null &
> [1] 1777
> speed# vmstat -m | grep USBdev ; sleep 1 ; vmstat -m | grep USBdev ; 
> sleep 1 ; vmstat -m | grep USBdev
>        USBdev    53     4K       -   270249  16,32,128,1024
>        USBdev    58     4K       -   271095  16,32,128,1024
>        USBdev    56     4K       -   272008  16,32,128,1024
> speed# vmstat -m | grep USBdev ; sleep 1 ; vmstat -m | grep USBdev ; 
> sleep 1 ; vmstat -m | grep USBdev
>        USBdev    54     4K       -   279649  16,32,128,1024
>        USBdev    57     4K       -   280544  16,32,128,1024
>        USBdev    54     4K       -   281423  16,32,128,1024
> 
> I don't know how relevant is the above, but it seemed strange,
> so I am posting it...

It's a normal case that in implementation it allocates a memory buffer
for transactions.  So if it increases without any reduction it'd be a
problem.

regards,
Weongyo Jeong




Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20090709023651.GA1553>