[comp.windows.x] Bizzare strangeness...the saga continues

adam@ncifcrf.gov (Adam W. Feigin) (01/12/90)

Well, I got nothing from my first posting, and I have a little more
information now, so I thought I'd try again. I built X11R4 on our
4/260C (cgtwo+gpone) earlier this week, and it was working for about 6
hours, then we lost the disk that the software was on. Sun came in and
replaced the disk AND the controller (to the latest rev), and I
reloaded all the software from tape, but it doesn't work anymore. The
message I get is:

xinit: Connection refused (errno 61): unable to connect to X server

I dont even get the nice stipple grey background on this machine, just
the error message. At first I thought perhaps the backup was
corrupted, so I rebuilt the entire distribution, but no luck. However,
we have another 4/260C (cgtwo, no gpone), and it works fine on that machine.

There's also another (bizzare) quirk. All of the clients work using
another machine's display, except for the fact that they have to have
the remote machines IP address:0  instead of hostname:0 as the display.
It's like gethostbyname is failing, because the client error message is:

Error: Can't Open display

(as opposed to Cant Open Display 'displayname')

This has me at my wits end....Does anyone have ANY suggestions as to
what my problem(s) may be, and how I might solve them ???

It's time to drink heavily......

							AWF




-- 
Internet: adam@ncifcrf.gov			Adam W. Feigin
UUCP: {backbonz}!ncifcrf!adam			Senior Systems Manager
Mail: P.O. Box B, Bldg 430	National Cancer Institute-Supercomputer Center
      Frederick, MD 21701		Frederick Cancer Research Facility

greg@sj.ate.slb.com (Greg Wageman) (01/16/90)

Opinions expressed are the responsibility of the author.

In article <1462@fcs280s.ncifcrf.gov> adam@ncifcrf.gov writes:
>Well, I got nothing from my first posting, and I have a little more
>information now, so I thought I'd try again. I built X11R4 on our
>4/260C (cgtwo+gpone) earlier this week, and it was working for about 6
>hours, then we lost the disk that the software was on. Sun came in and
>replaced the disk AND the controller (to the latest rev), and I
>reloaded all the software from tape, but it doesn't work anymore. The
>message I get is:
>
>xinit: Connection refused (errno 61): unable to connect to X server
>
>There's also another (bizzare) quirk. All of the clients work using
>another machine's display, except for the fact that they have to have
>the remote machines IP address:0  instead of hostname:0 as the display.
>It's like gethostbyname is failing, because the client error message is:
>
>Error: Can't Open display
>
>(as opposed to Cant Open Display 'displayname')

You don't possibly have a mixed network of 4.0.x and 3.x machines, do
you?  With multiple yp-servers?  I have seen 4.0.x machines hang
trying to do yellow-pages lookups when bound to pre-4.0.x servers.

If you have another yp server, try doing a "ypset" to it, and see if
the problem clears up.  Oh, and you will need to run ypbind with the
undocumented "-ypset" option to allow this if you're running 4.0.x.

(For all concerned, it would really help in getting answers to
questions if you would say what OS you are running!)

Copyright 1990 Greg Wageman	DOMAIN: greg@sj.ate.slb.com
Schlumberger Technologies	UUCP:   {uunet,decwrl,amdahl}!sjsca4!greg
San Jose, CA 95110-1397		BIX: gwage  CIS: 74016,352  GEnie: G.WAGEMAN
 Permission is granted for reproduction provided this notice is maintained.

earle@POSEUR.JPL.NASA.GOV (Greg Earle - Sun JPL on-site Software Support) (01/16/90)

The solution to Adam Feigin's problem was that there was an entry in his
/etc/X0.hosts file that caused ResetHosts() (in server/os/4.2bsd/access.c)
to get indigestion, and when the host was to be added (via NewHost()),
something was screwed up and made the acopy() call (macro-ized bcopy()) blow
the server out of the water.  Thus the ECONNREFUSED - there was no server
there to listen on the UNIX domain socket by then (/tmp/.X11-unix/X0).

By moving his /etc/X0.hosts file out of the way, the server started.  Adam is
still trying to determine what exactly is in the /etc/X0.hosts file that caused
the NewHost() routine to blow its cookies.

Note to the general X Universe: is there a particular format that /etc/Xn.hosts
has to follow?  Any requirements for particular whitespace (like TABs), etc. ??

--