home *** CD-ROM | disk | FTP | other *** search
- Path: sparky!uunet!charon.amdahl.com!pacbell.com!att!linac!pacific.mps.ohio-state.edu!cis.ohio-state.edu!zaphod.mps.ohio-state.edu!cs.utexas.edu!sun-barr!news2me.ebay.sun.com!exodus.Eng.Sun.COM!sun!amdcad!angelo!shah
- From: shah@angelo.amd.com (Amit Shah)
- Newsgroups: comp.dcom.cell-relay
- Subject: Re: >>>>Future of IP routers
- Message-ID: <1992Aug30.194536.2545@amd.com>
- Date: 30 Aug 92 19:45:36 GMT
- References: <1992Aug28.164008.4015@sics.se> <p6dueak@rhyolite.wpd.sgi.com>
- Sender: usenet@amd.com (NetNews)
- Organization: Advanced Micro Devices, Santa Clara, CA
- Lines: 38
- Nntp-Posting-Host: angelo
-
- In article <p6dueak@rhyolite.wpd.sgi.com> vjs@rhyolite.wpd.sgi.com (Vernon Schryver) writes:
- >In article <1992Aug28.164008.4015@sics.se>, craig@sics.se (Craig Partridge) writes:
- >> ...
- >> (any FDDI throughput test over a real ring that doesn't mention the TTRT
- >> that the testers set for the network is worthless) ...
- >
- >While I agree with the rest of Craig's article, I disagree with this
- >caveat. FDDI TTRT by itself does not matter. What matters is how
- >often you get to transmit and consequently how much of the time the
- >ring is idle. Unless you are using TCP Large Windows (as Cray does and
- >we all should), legal TTRT's do not crimp your style. FDDI is less
- >than 12.5KBytes/millisecond, the maximum standard TCP window is only
- >64KBytes, and the minimum legal T_Req is on the order of 4 ms.
- >
- I have a slight disagreement with your statement that FDDI TTRT by itself
- does not matter. It depends on the type of traffic you put on the network.
- If you are talking about bursting traffic in excess of 100 Mbps (such as
- high-end imaging), with >2 stations on the ring then the
- TTRT does matter (Lower TTRT will have lower throughput and buffer overflows).
-
- However, if you are talking about normal traffic (burst rates not exceeding
- 100 Mbps) then you are correct within a few percent points.
-
- >No interesting maximum-speed FDDI benchmark is going to be run on
- >anything but a practically idle FDDI ring. Very few FDDI rings have
- >physical latencies large enough to matter, so that how frequently you
- >must release the token mostly affects how frequently you need, expect,
- >or hope to get TCP ACK's.
- >
- >
- >Vernon Schryver, vjs@sgi.com
-
-
- --
- Amit Shah
- Sys. Engg. & Prod. Planning Engineer Ph.:(408)987 2479
- High Speed Networking Fax:(408)987 2814
- 901 Thompson Place, M/S 70 email:amit.shah@amd.com
-