home *** CD-ROM | disk | FTP | other *** search
- Path: sparky!uunet!zaphod.mps.ohio-state.edu!saimiri.primate.wisc.edu!ames!nsisrv!Pt!wainner
- From: wainner@Pt.hq.af.mil (Scott Wainner)
- Newsgroups: comp.dcom.cell-relay
- Subject: Re: Computers dont like ATM?
- Message-ID: <wainner.722408411@hq>
- Date: 22 Nov 92 05:00:11 GMT
- References: <1992Nov18.025754.14749@trl.oz.au>
- Organization: Headquarters Air Force, The Pentagon Washington, DC
- Lines: 49
-
- lampard@titan.trl.OZ.AU (Greg Lampard) writes:
-
- > Paul Green statements: he mentioned that 53 byte cells are too small
- >for computers (or words to that effect). His claim was that small
- >packets are good for delay-sensitive applications, but they make a lot
- >of unnecessary work for the computer that has to receive them, since it
- >has to process a header every 53 bytes, and often just stores the cell
- >to build up a larger packet for the next layer up the protocol stack.
-
- It is true that 53 bytes is a nuisance to the computer. The computer
- industry proposed a 69 byte cell (ie 64 byte payload) but a compromise
- was made with Europe which wanted a 32 byte cell. So, we get a 44 or
- 48 byte payload depending on the AAL used. I think 48 seems logical
- since it's a multiple of 8 bytes. The reason for the fixed cell size is
- to minimize buffer requirements in the switch. Since all the cells are
- the same length, the switch can process the cells in "parallel". The
- other benefit of fixed size cells is that it may be implemented in
- hardware that will perform frame reconstruction must faster than
- software processes.
-
- >He also seemed to be suggesting that this storage and retrieval process
- >could become a bottleneck (at 100s of Mbit/s?) because of the limitations
- >of memory bandwidth. He was advocating variable length packets (an idea
- >called PTM - packet transfer mode - that IBM has proposed), because it
- >puts the complexity in the switch rather than in the receiving
- >terminal.
-
- Exactly, we're trying to make the switch faster by eliminating the
- protocol processing and implementing the switching in hardware.
-
- >From this, his conclusion was (I think) that ATM is good for switching
- >voice traffic, but not so good for computer networks. This seems to go
- >against what's actually happening, with people building ATM LANs while
- >telcos hang back, not sure if ATM is the right way to go.
-
- I think ATM has great merits but I'm not quite sure how fault tolerance
- will be introduced in the network since we're using VCIs and VPIs to
- perform the routing. Is this static once the call setup is complete?
- How does the switch know to reroute a failed connection without forcing
- the application to "reopen" the path.
-
-
- >Greg.
- >g.lampard@trl.oz.au | communications engineer.
-
- Scott Wainner
- swainner@tasc.com.
- Communications Systems Engineer
-
-