[Mono-list] Why is MONO network streaming slow?

[Alan McGovern]

Andreas, could you use windows-mono and run the the MS.NET test and Mono
test on the same OS and same machines? That's the only way to get a clear
indication if there is actually a problem with mono or not.

Comparing linux-mono with windows MS.NET has far too many variables to even
know where to begin. For example, you're linux box could just have terrible
drivers whereas the MS box has very good drivers. That alone could be the
cause of the massive difference you're seeing.

Alan.

On 6/19/07, William Huskey <william.huskey at gmail.com> wrote:
>
> sorry, the issue has more to do with how many other systems are on the
> network, are they the same computer and NIC. if not it may be the NIC driver
> or the Linux TCP/IP stack first. I can get 11MB second on 100 Mb also using
> NFS or SMB but its not until the first 20 - 30 meg is sent that it is seen.
> in your test I see you do not have buffering turned off -D which can throw
> off a measurement.
>
> Overhead on a network can be
> packet re-packaging (going through NAT)
> Packet handling as in routing
> TCP to TCP handshaking and resend
>
> also packet size will produce a difference in speeds between system.
> packages up to 1024 bytes in size should transfer the same speed (should!!)
>
> MTU size also effects most of the transfer speed and is set by the TCP/IP
> stack not the program and is specific to the OS.
> http://www.cisco.com/univercd/cc/td/doc/product/vpn/client/4_6/admin/vcach8.htmwill give information on how to set that.
>
> to know if this is a MONO problem we need to have it tested where each
> machine is the same build
> test with a Switch in place (25 - 50 test no buffering)
> test with crossover (this will tell if the network cards them selves have
> issues)
>
> use a independent machine to also capture available bandwidth on each
> side.
>
> at work we use Chariot
> http://www.netiq.com/news/releases/release.asp?cid=20020513100815ECVG to
> test this.
>
>
>
>
> On 6/19/07, Alan McGovern <alan.mcgovern at gmail.com> wrote:
>
> >
> >
> > On 6/19/07, Adam Tauno Williams < adam at morrison-ind.com> wrote:
> > >
> > > > wow there is so much wrong with this test,
> > > > First I find it hard to believe that you got up to 90% efficiency on
> > > a
> > >
> > > I noticed that too.
> > > "NET --> .NET : 94 Mbit/s   O.K." on a 100Mb/s network.
> > > Bogus.  No way.
> > >
> > > >  transfer  at all, network over head should have knocked down your
> > > > transfer more than that.
> > >
> > > Yep.
> >
> >
> > Nope,  i've transferred files  on my home network and reached over
> > 90Mbit/sec on a 10/100 network. It's definitely not beyond the realms of
> > possibility that he achieved 93Mbit/sec.
> >
> >
> >
> > _______________________________________________
> > Mono-list maillist  -  Mono-list at lists.ximian.com
> > http://lists.ximian.com/mailman/listinfo/mono-list
> >
> >
>
>
> --
> William S. Huskey
>
> UNIX Systems Engineer
>
> Maritime Systems Solutions Division
>
> SAIC
>
> Tel: 425-267-5643
>
> Email: huskeyw at saic.com
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.ximian.com/pipermail/mono-list/attachments/20070619/da806619/attachment.html