AWalker@RED.RUTGERS.EDU (*Hobbit*) (01/28/86)
I have a vax on an ethernet cable with a few other DECNET vaxen on it also. In particular, I'm host 13, the others are 2 [NB] and 10 [CHEMA]. NB routes to a bunch of other hosts via a DMC or something. In farting around with tasks and things, I've run across entries in netserver.log of the type '' connect request received from 1026::"0=USER" which turns out to be host 2. I.e. if I do SET HOST 1026 from my host, I get to host 2. My host is nonrouting, has no knowledge of host 2, 1026, or anything like that in its own tables, so it doesn't know what to call the incoming host. Do host numbers wrap around at 1024, is 2 translating to 1026, and why would my host be adding extra bits to the incoming host number? In other words, what's going on? _H* -------
SYSTEM@CRNLNS.BITNET (02/02/86)
Subj: re: Another interesting one > >... I've run across entries in netserver.log >of the type '' connect request received from 1026::"0=USER" which turns >out to be host 2. ... > >_H* All DECnet node numbers are actually a 16 bit number. The low order 10 bits are allocated to the node number within the area (0-1023 decimal). The high order 6 bits indicate the area number (0-61 decimal). Unless you specify otherwise, under VMS v4.x all DECnet nodes are in area 1. If you don't have a DECnet node in your local network database, its node name will show up in log files as the area number multiplied by 1024 added to the node number. I trust that this explains what you are seeing. Selden Ball Cornell University, LNS NYNEX: (607) 256-4882 Wilson Synchrotron Lab BITNET: SYSTEM@CRNLNS Judd Falls & Dryden Road ARPA: SYSTEM%CRNLNS.BITNET@WISCVM.WISC.EDU Ithaca, NY 14853 (SYSTEM%CRNLNS.BITNET@WISCVM.ARPA) PHYSnet/HEPnet: LNS61::SYSTEM (node 43.251/44283)