There is work going on both to improve TCP so it works better on Fast Long Distance networks (see for example http://www.slac.stanford.edu/cgi-wrap/getdoc/slac- pub-10402.pdf for a comparison of the performances of some of these new versions of TCP, and to come up with replacements such as SCTP and UDT (see http://www.ietf.org/internet-drafts/draft-gg-udt-0 1.txt) which is a reliable transport protocol built on top of UDP.
We did make tests with UDT on the SC2004 testbed and it perfomed very well (~4.5Gbits/s) which was not as good as we saw with TCP but it is early days for the UDT so it is looking hopeful.
In answer to the question: "Nowhere in the article does it say how long they ran the test for. A second? A minute? An hour?"
On many of the 10Gbps paths we were able to sustain sending over 99% of the available bandwidth for hours at a time. We sustained an aggregate of over 100Gbits/s for about 2 minutes. The median aggregate bandwidth over 48 minutes was about 66Gbits/s. The HEP bandwidth challenge test ran for about 48 minutes.
Much more information is available at: http://www-iepm.slac.stanford.edu/monitoring/bulk/ sc2004/hiperf.html
Slashdot Mirror
There is work going on both to improve TCP so it works better on Fast Long Distance networks (see for example http://www.slac.stanford.edu/cgi-wrap/getdoc/slac- pub-10402.pdf for a comparison of the performances of some of these new versions of TCP, and to come up with replacements such as SCTP and UDT (see http://www.ietf.org/internet-drafts/draft-gg-udt-0 1.txt) which is a reliable transport protocol built on top of UDP.
We did make tests with UDT on the SC2004 testbed and it perfomed very well (~4.5Gbits/s) which was not as good as we saw with TCP but it is early days for the UDT so it is looking hopeful.
In answer to the question: "Nowhere in the article does it say how long they ran the test for. A second? A minute? An hour?"
/ sc2004/hiperf.html
On many of the 10Gbps paths we were able to sustain sending over 99% of the available bandwidth for hours at a time. We sustained an aggregate of over 100Gbits/s for about 2 minutes. The median aggregate bandwidth over 48 minutes was about 66Gbits/s. The HEP bandwidth challenge test ran for about 48 minutes.
Much more information is available at: http://www-iepm.slac.stanford.edu/monitoring/bulk