Hallo. I recently bought an Apollo Domain coputer for 15 dollars.
It is an Domain Series 3500 with 4 meg of ram, tape drive, 170 ESDI
hard disk and a 19" color monitor.
No manuals, no SR tapes, configured with Domain/IX Aegis 9.7.2.1 and
BADGE CAD program.

I tried to upgrade this machine with Domain/OS SR 10.3 and ended with
complete failure. Not knowing the concept of AA installation, I
executed in DM 9.7.2.1. the doman_os SR 10.3. It began to ugrade the
system, but informed me to late about the incompatible disk format of
9.7.2.1. and  SR 10.3. Left with an unbootable disk and no way to
return, I started a complete installation in Service Mode.
Executed INVOL, CALENDAR, DOMAIN_OS, GO, MINST, Aegis Small
installation and ended with MINST COMPLETE, check your log file for
errors and reboot. Sure !!. 
System reboots in normal mode and asks me for a login id and passwd. 
Yes , what now ? Finally figured out that the only way to enter the
system was LOGIN : user, Passwd : <enter>. Then it began to complain.
Network registry not available, using local registry. Registries
unavaileble !. No doubt about it I NEEDED A MANUAL. Got one and
figured out that this was normal. Further went my quest for knowledge.
Executed  MANUAL instructions 
//node_18b67/install/toold/rgy_create
/etc/server -p  /etc/llbd &      <ended wikth server error, no such
file or directory>
crf /etc/daemons/llbd  /etc/daemons/rgyd
/etc/server -p  /etc/glbd -first -create &  < same error>
crf /etc/daemons/glbd
 and reboot and login as root, passwd apollo.

No way, this didn't seem to work.  Changed the instructions to :
/etc/server -p /etc/ncs/llbd &  < worked>
crf /etc/daemons/llbd /etc/damons/ncs/rgyd < worked.
/etc/server -p /etc/ncs/glbd -first -create & < worked>
crf /etc/daemons/glbd  < worked >

rebooted and ended with the same problem: No network registry
available. Using local registry. Registried unavailable.
Login as user.none.none  Presets like user:< root > and
 passwd:< apollo > do not work.

I have no clue how to resolve this problem with the registries. The
manual  is not correct, but I must be doing something wrong to.

Who can help me.

Pierre ( lorier@worldonline.nl).

lorier@worldonline.nl (Pierre Lorier) writes:

>Hallo. I recently bought an Apollo Domain coputer for 15 dollars.  >It is
an Domain Series 3500 with 4 meg of ram, tape drive, 170 ESDI >hard disk
and a 19" color monitor.  >No manuals, no SR tapes, configured with
Domain/IX Aegis 9.7.2.1 and >BADGE CAD program. 

Snip >>>>>>>>>>>>>>>>>

Here is a short letter for anyone in need of moving or setting up a
Registry. There may be errors so  use at your own risk.
Comments or corrections welcome. 

Jerry Wright
JLC inc.





To move an apollo from a network and make it a stand alone 
node do the following. With out reloading the "OS".
You must be "root" .  Some of these files and daemons will not be
on all machines. 

 kill            llbd
                 glbd
                 rgyd
                    
Check for these files and "mv" them to ".old" or whatever if
they are there.

 /sys/node_data  glb.e       (glb.e.bak)
                 glb.p
                 hint_file
                 ipc_data
 /etc/ncs        glb_obj.txt
                 glb_site.txt
                               
Remove everything in 
 /sys/registry              (keep the dir) All the data and dir's 
                             will be remade.   


There are a lot of links to other nodes. Some are the "man pages.
/sys/help, many of the /usr/spool dir., lp, uucp, ect. These all need to 
be copied before hand. In /etc the host,networks type files are all plain
text files. Most will be linked to another node. The ownership of many of
the spool files are important. LP must be owned by LP etc, so watch what you
copy and who owns them. 


Make sure You have a copy of rgy_create. This is in the "aa"
under tools. 


At this point, It is best to remove the node from the ring if it 
is going to be a stand alone machine. Use the "netsvc" command to 
disable the port. See man pages or unhook it.   
             
Run rgy_create


For 10.3.x

  /etc/server -p /etc/ncs/llbd &
  /etc/server -p /etc/ncs/glbd -create -first -family dds &
  /etc/server -p /etc/rgyd &


For 10.4

  /etc/server -p /etc/ncs/rpcd  &  (See man pages)
  /etc/server -p /etc/ncs/glbd -create -first -family dds &   (see man pages)
  /etc/server -p /etc/rgyd &
                                           
Do a "ps" command to make sure these are running. 

Run /etc/edrgy -v and make sure the default users are there.
    Do not change anything!  


Make sure these file are in /etc/daemons. It is best not to start anything 
else untill you are sure it will boot "OK". So do a "ls" on /etc/daemons.

 /bin/touch   /etc/daemons/rpcd (for 10.4.x)
              /etc/daemons/glbd
              /etc/daemons/llbd (for 10.3.x) 
              /etc/daemons/rgyd
 
For TCP-IP you will have to check and edit the following.

 /etc/   rc.local
         hosts                               
         hosts.equiv
         networks

These file are used during boot to config the node so make sure they are
 correct.
/etc/rc* are a link to /sys/node_data/etc/rc* and there may be other copies of
these there in .bak or .orig form. Another way to get a clean set of /sys/node_
data files is to boot a machine diskless. This will create a /sys/node_data.node.id
dir. on the partner node. Where "node.id" is the id of the node you booted.  

 /etc/    rc  
          rc.local
          rc.user
          profile (make sure there is nothing here that will cause a problem)


reboot the node. 

If this node is being moved to another ring, Then you only have to have
llbd or rpcd in the daemons dir. Then reboot the node. The site rgy will 
build the local rgy files in a short period. Do not start the other daemons!
One final note when moving nodes. Make sure the Netid is correct. Use "netsvc"
to check this Before rebooting! This is only needed if the node was on a
different ring. If the netid is going to change, Then /etc/host and /sys/
node_data/etc/rc.local need to be changed at least. With the wrong Netid
no other nodes on that ring will see it.          

>>> Standalone <<<<

If the node has unresolved link's in the areas it needs on boot, it will 
hang for long periods. If the token-ring port is disabled or looped back
this will help.  The "netsvc" command can be put in the "rc.local" file
to disable the network port on boot..
       
I'am not testing the above as I write It. So there may be a few spelling and 
typo errors. This is a very brief listing of what needs to be done at this 
level. Many  files need to be changed to help in the login process and of
coarse all the links that may be pointing to the old net_work.  

Jerry

Pierre Lorier (lorier@worldonline.nl) wrote:
: Hallo. I recently bought an Apollo Domain coputer for 15 dollars.
: It is an Domain Series 3500 with 4 meg of ram, tape drive, 170 ESDI
: hard disk and a 19" color monitor.
: No manuals, no SR tapes, configured with Domain/IX Aegis 9.7.2.1 and
: BADGE CAD program.

    <- snip ->

:  and reboot and login as root, passwd apollo.

    <- snip ->

: Login as user.none.none  Presets like user:< root > and
:  passwd:< apollo > do not work. 
: I have no clue how to resolve this problem with the registries. The
: manual  is not correct, but I must be doing something wrong to. 
: Who can help me.


    Only one mistake (I think) the password on a newly installed system is 
    -apollo-        the "-" are part of the password.
    
    


--
Alan Massey, Computer Systems Manager, Philips Semiconductors, Southampton, UK
email : massey@ukpsshp1.serigate.philips.nl    (SERI : massey@ukpsshp1)
Tel   : +44 (0)1703 316450                      Fax : +44 (0)1703 316305