home *** CD-ROM | disk | FTP | other *** search
- Newsgroups: comp.protocols.tcp-ip.ibmpc
- Path: sparky!uunet!utcsri!torn!watserv2.uwaterloo.ca!watserv1!demorgan.uwaterloo.ca!erick
- From: erick@demorgan.uwaterloo.ca (Erick Engelke)
- Subject: Re: Fragmented IP packets: any PD implementations?
- Message-ID: <BxGtps.BHM@watserv1.uwaterloo.ca>
- Keywords: IP fragments TCP/IP
- Sender: news@watserv1.uwaterloo.ca
- Organization: University of Waterloo
- References: <1992Nov8.203621.6902@mnemosyne.cs.du.edu>
- Date: Mon, 9 Nov 1992 20:17:51 GMT
- Lines: 60
-
- > dpayne@nyx.cs.du.edu (Dirk Payne) writes:
- >I have reviewed WATTCP, PCIP, and both CUTCP and NCSA's versions of telnet
- >but none of these implemetations handle IP fragments. In order to be a "real"
- >TCP/IP package (connected to the "real" world) fragments cannot be ejected
- >out of hand.
-
- The question I have is why you are needing frags. There are only a
- few situations which should generate fragments in a user's PC, large
- UDP packets and incorrectly configured TCP systems. (Frags in
- gateways are quite a different story.)
-
- First, let's dispell the TCP case. Your user's PCs are located on
- an Ethernet, Arcnet, TR, slip line, or whatever. Setting the correct
- tcp segment size for your hardware will guarentee that you and your
- TCP peers will never attempt to send a packet larger than that segment
- size and so there will be no fragments. And systems like WATTCP
- and CUTCP use Van Jacobson's adaptive window, so they will offer
- fast bulky TCP transfers without requiring frags.
-
- A common misconception is that we will get fragmented packets from
- outside our LAN, such as gateways through slower media. But it is
- the gateway's responsibility to reassemble the frags before putting
- the data onto the LAN at the other end. That's why people using
- all these freeware TCPs didn't know they didn't have frag support.
-
- Some TCPs, like FTP's PC/TCP, actually determine the largest segment
- size which will not fragment along the path. It's coverred in
- the RFC's and quite a clever idea worth reading.
-
- The remaining use of fragments is for UDP based services such as
- NFS/RPC/UDP or possibly SNMP. I have always been critical of the
- use of fragmented RPC's and wished people would migrate over to TCP
- as they are starting to do now for NFS.
-
- Here's my frustration, fragments rely on 100% delivery and are
- costly in gateways. And when you encounter problems on your LAN,
- large UDP packets clog it in a needless and futile manner whereas
- TCP adapts.
-
- I might be convinced to add frag support, but would appreciate
- opinion if it is a good idea, in particular, which applications
- could use it. NFS is not a great answer since it would be useful
- even with 1K packet sizes, the problem there is the lack of an
- implementation. I know the Host Requirements RFC tells us what
- to do, but I would prefer to see an application which needs it.
-
-
- >Has anyone implemented IP fragment handling in any of the above mentioned
- >products? My personal favorite is WATTCP, but I would be willing to entertain
- >any PD implementation that handles frags. My only alternative is to roll my
- >own :-(.
-
- Like I said, gateways are a different matter. Phil Karn's NOS does
- frags/reassembly, but it isn't PD.
-
- Erick
- --
- ----------------------------------------------------------------------------
- Erick Engelke WATTCP Architect
- erick@development.uwaterloo.ca TCP/IP was easy but i still can't work VI
-