[Click] high throughput hardware configurations
Eddie Kohler
kohler at CS.UCLA.EDU
Sun Feb 27 22:31:05 EST 2005
Hi Abdul,
Click is a forwarding path -- it doesn't have a lot of support for transport
protocols. In fact it has no support at all for SCTP.
Eddie
Abdul Hakeem wrote:
> Hello,
> I wanted to know if it's possible to compile Click with SCTP options.
> Best regards,
> Abdul Hakeem
> -----Original Message-----
> From: click-bounces at amsterdam.lcs.mit.edu
> [mailto:click-bounces at amsterdam.lcs.mit.edu] On Behalf Of Nikitas Liogkas
> Sent: 19 February 2005 02:27
> To: click at amsterdam.lcs.mit.edu
> Subject: [Click] high throughput hardware configurations
>
>
> Hello Clickers!
>
> we are trying to put together a testbed for running Click experiments here
> at UCLA, and I was wondering whether you could share with us the hardware
> configurations (processor, PCI bus, main memory, motherboard, network cards,
>
> Linux kernel and Click version) that worked for you, and that resulted in
> high throughput with Click.
>
> For example, I know that Giorgio Calarco from the University of Bologna,
> Italy, has been using the following kind of hardware to achieve loss-free
> throughput of 737 Kpackets/sec running uniprocessor Click:
> - dual Intel Xeon 2.80 GHz (with HyperThreading disabled)
> - Intel E7501 chipset motherboard with a 533 MHz FSB and four PCI-X slots
> - four Intel 88544EI Gigabit ethernet cards, two of them plugged in
> 64-bit/133MHz PCI-X slots, and two in 64-bit/100 MHz PCI-X slots
> - main memory?
>
> We'd love to know what configs worked for you, and any suggestions for
> hardware purchases you might have. Thanks so much!
>
> nikitas
> _______________________________________________
> click mailing list
> click at amsterdam.lcs.mit.edu
> https://amsterdam.lcs.mit.edu/mailman/listinfo/click
>
> _______________________________________________
> click mailing list
> click at amsterdam.lcs.mit.edu
> https://amsterdam.lcs.mit.edu/mailman/listinfo/click
More information about the click
mailing list