home *** CD-ROM | disk | FTP | other *** search
- Path: sparky!uunet!think.com!barmar
- From: barmar@think.com (Barry Margolin)
- Newsgroups: comp.unix.wizards
- Subject: Re: Please explain...
- Date: 21 Nov 1992 02:36:47 GMT
- Organization: Thinking Machines Corporation, Cambridge MA, USA
- Lines: 38
- Message-ID: <1ek7bvINNj50@early-bird.think.com>
- References: <1992Nov18.190632.13873@news.uit.no> <1992Nov20.093151.19989@thunder.mcrcim.mcgill.edu>
- NNTP-Posting-Host: telecaster.think.com
- Keywords: Telnet, pseudo-terminals, sockets
-
- In article <1992Nov20.093151.19989@thunder.mcrcim.mcgill.edu> mouse@thunder.mcrcim.mcgill.edu (der Mouse) writes:
- >In article <1992Nov18.190632.13873@news.uit.no>, orjanr@staff.cs.uit.no (Xrjan Robertsen) writes:
- >> I thought all communication had to pass through telnetd, but
- >> monitoring this process (with monitor) shows no I/O activity when
- >> telnet and the remote shell is communicating. How can this be ?
- >
- >monitor, whatever that is,
-
- My guess is it's something similar to trace or truss.
-
- > is either broken or is not designed to do
- >what you're trying to do with it. If the remote system (the one where
- >telnetd and the remote shell are running) is even remotely "normal" (by
- >which I really mean something more like "typical"), all communication
- >between the net and the remote shell must go through telnetd.
-
- Unfortunately, Mr. Robertsen didn't mention what kind of system he's using.
- Perhaps it has a STREAMS module that permits two streams devices to be
- connected directly, so that most data doesn't have to go through the
- user-mode process. The ITS operating system for PDP-10's has a system call
- that links two devices in this manner; it can be given a set of characters
- that should cause a break in this automatic forwarding, e.g. so that the
- daemon process can handle special characters such as TELNET IACs or
- terminal escape sequences.
-
- Another possibility is that the system has the telnet server built into the
- kernel (much as many systems have NFS servers in the kernel). In that
- case, telnetd only exists to make the system call that starts it up
- (analogous to nfsd).
-
- Now, I don't personally know of any Unix systems that do either of these,
- but I don't claim to be familiar with many obscure Unix variants (oops,
- does that mean I'm not wizardly enough to post this? :-)
- --
- Barry Margolin
- System Manager, Thinking Machines Corp.
-
- barmar@think.com {uunet,harvard}!think!barmar
-