csvchris@jarthur.Claremont.EDU (Chris Yoder) (10/31/89)
Well, at long last (a year and a half and two TCP/IP's later) I've got ANU
News up and running on my workstation on the cluster. Software: VMS 5.2, VAX C
3.0, News 5.8a (with Multinet patches from Joe Deck at Wesleyan). All groups
are served off of a local Un*x machine (a Sequent). I have dedicated an old
RA81 on the cluster boot node (an 8600) to be my news disk (for caching and the
NEWS_MANAGER: and NEWS_ROOT: directories... better safe than sorry).
I've let the rest of the system's people loose on playing with News, and
so far the responses have ranged from rabid enthusiasm to ``it's better than
logging onto a Un*x machine and using rn''... However, I'm having a problem
with posting from the VMS side, and I also want to know what my NEWS.SYS file
should look like so that all systems in the cluster can use the same file (and
be known by their own DECnet names).
Here's a copy of the uncommented portions of NEWS.SYS:
#
echmc:alt,bionet,biz,ca,comp,ddn,gnu,la,misc,news,oc,pnw,pubnet,\
rec,sci,soc,talk,u3b,unix-pc,mudd,to.echmc,local,junk,control::
#
jarthur:alt,bionet,biz,ca,comp,ddn,gnu,la,misc,news,oc,pnw,pubnet,\
rec,sci,soc,talk,u3b,unix-pc,mudd,to.jarthur::
#
echmc is my workstation (a VS 3100), and jarthur is the Sequent. Right now we
have 13 nodes in the cluster (no CI hardware). I was thinking of modifying
this to read:
#
echmc:alt,bionet,biz,ca,comp,ddn,gnu,la,misc,news,oc,pnw,pubnet,\
rec,sci,soc,talk,u3b,unix-pc,mudd,to.echmc,local,junk,control::
#
hmcvax:alt,bionet,biz,ca,comp,ddn,gnu,la,misc,news,oc,pnw,pubnet,\
rec,sci,soc,talk,u3b,unix-pc,mudd,to.hmcvax,local,junk,control::
#
ymir:alt,bionet,biz,ca,comp,ddn,gnu,la,misc,news,oc,pnw,pubnet,\
rec,sci,soc,talk,u3b,unix-pc,mudd,to.hmcvax,local,junk,control::
#
# Other nodes in similar fashion.
#
jarthur:alt,bionet,biz,ca,comp,ddn,gnu,la,misc,news,oc,pnw,pubnet,\
rec,sci,soc,talk,u3b,unix-pc,mudd,to.jarthur::
#
Is this the correct way of doing things in a cluster? Is there something
obviously wrong with this file that I need to fix so that I can post? Can I
make this NEWS.SYS even more general? (I want everything that jarthur gets to
come up on all other nodes.) One thing that I don't want is for every message
posted on the cluster to have their return address set to my workstation (where
the NEWSSKIM runs nightly).
Oh, and is it reasonable to expect a newsgroup creation run of over 78
hours with this kind of distribution? (It seems long to me.)
-- Chris Yoder Internet --- Chris@HMCVAX.Claremont.Edu
Harvey Mudd College Bitnet ----- Chris@HMCVAX.Bitnet