Trying to outsmart AUTOGEN and CLUSTER_CONFIG is never a good idea.
After the disk image transfer, you are no longer running a 6220, but a
3900. After the disk image transfer and first (minimal) boot, you should
AUTOGEN and, as you found out, NETCONFIG. Then your cluster boot master
should be OK.
Next, remove the VAX4000 from the cluster using CLUSTER_CONFIG, and
re-add it using the same procedure. This ensures all (device) pointer
and names are OK. DSVCONFIG should not be needed for the cluster
satellite loads, only for serving DECserver image loads.
*Wilm*
Daniel Parrott schreef op 20-6-2016 om 22:16:
I have two physical systems, a VAX 6220 and a VAX 4000-300. The latter
system is set up to do a network bootstrap off the 6220 using an EZA0
network device.
First, using vddriver and backup/image, I generated and retrieved disk
images of the RA92 drives from the 6220. With these disk images, and
specifying MicroVAX 3900 on the SimH simulator, I am able to boot up
an emulated version of the 6220 on my laptop. So far so good.
My first goal is to get the physical VAX 4000-300 to do its network
bootstrap off the emulated 6220 rather than the physical 6220. By the
way, I am running SimH 4.0 on Windows 8, it seems to work OK.
The first problem I ran into was that the simulator presents a
different network adapter to the emulated 6220 than it is accustomed
to seeing. It typically sees a "BNA-0" interface, but the simulator
uses "QNA-0", as derived from the XQA0 device. So, I found out that I
needed to run netconfig.com. When I ran that script, it brought the
circuit online.
But when I tried to do network boot of the physical VAX 4000-300 off
the emulated 6220, it kept saying 'retrying network bootstrap'. So
evidently that was not sufficient. Then I came to discover that the
6220 uses DECserver, and there is a dsvconfig.com script. Once again,
because the simulator presented a different network device than the
6220 was accustomed to seeing, the configuration had to be updated.
So okay, I ran the dsvconfig.com script, and made it aware of the fact
that it is now a "QNA-0" device rather than a "BNA-0" device.
So after running both netconfig.com and dsvconfig.com on the emulated
6220, the physical 4000-300 no longer says "retrying network
bootstrap". Instead it now appears to be able to communicate, but it
says "unable to locate boot file":
>>> b eza0
(BOOT/R5:0 EZA0)
2..
-EZA0
1..0..
%VAXcluster-I-SYSLOAD, system loaded from node ORION (AA-00-04-00-B6-11)
%VAXcluster-F-BADOPEN, unable to locate BOOT file
?06 HLT INST
PC = 00007395
>>>
Using Wireshark, I captured some of the packets to try to glean some
more details, and I see messages like "VAXcluster-F-NOSUCHDEV,
nonexistent drive" and "VAXcluster-F-READERR, I/O error reading boot
file".
I am really stumped on this one because basically the disk images
should be one-to-one copies of what is on the physical VAX 6220, so
other than having to tell it that it has a different network adapter
(QNA-0 rather than BNA-0), why shouldn't it just work? Does SimH not
support MOP or DECserver? I did check, and the sysboot.exe file is
present in dua0:[sys14.syscommon.sysexe], and sys14 is the directory
for the VAX 4000-300 to use for its network boot. So I believe the
boot file is available.
By the way, the VAX 6220 is running VMS 5.5-2H. And I have set
VAXCLUSTER to 2.
Oh and one more thing, when the 4000-300 fails to boot, on the
simulator side of things I see the following:
%%%%%%%%%%%% OPCOM 20-JUN-2016 15:10:06.58 %%%%%%%%
Message from user DECNET on ORION
DECnet event 0.3, automatic line service
From node 4.438 (ORION), 20-JUN-2016 15:10:06.58
Circuit QNA-0, Load, Requested, Node = 4.764 (GUINAN)
File = DISK$VMSRL5:<SYS14.>, Operating system, Ethernet address =
08-00-2B-23-9A-F4
%%%%%%%%%%%% OPCOM 20-JUN-2016 15:10:11.58 %%%%%%%%
Message from user DECNET on ORION
DECnet event 0.3, automatic line service
From node 4.438 (ORION), 20-JUN-2016 15:10:06.75
Circuit QNA-0, Load, Successful, Node = 4.764 (GUINAN)
File = DISK$VMSRL5:<SYS14.>, Operating system, Ethernet address =
08-00-2B-23-9A-F4
%PEA0, Port has Closed Virtual Circuit - REMOTE NODE GUINAN
So it almost looks like it has done a successful load of the system
software to the 4000-300, but then the port closes presumably due to
not being able to locate the boot file.
My thoughts are that maybe:
1) there is some other config script that needs to be run when using
disk images from one VAX system (physical 6220) to another (simulated
MicroVAX 3900)
2) SimH on Windows 8 does not really support MOP or DECserver stuff?
3) there is some permissions issue with the boot file (sysboot.exe?),
but since it is off a disk image (which should be a one-to-one exact
copy), how so?
Any ideas? Thanks in advance.
_______________________________________________
Simh mailing list
[email protected]
http://mailman.trailing-edge.com/mailman/listinfo/simh
_______________________________________________
Simh mailing list
[email protected]
http://mailman.trailing-edge.com/mailman/listinfo/simh
