12 Jun
2015
12 Jun
'15
7:01 a.m.
Using a version 58 console tape image provided to me by one list member, and
massaged into a usable state by another list member, I just booted
OpenVMS 7.3 off the R80 drive on my VAX-11/730 for the first time since
buying the machine. Woohoo!
Excellent!
For some reason, I was unable to do that with the version 57 images that I
downloaded off the net. Maybe there's something wrong with the VMB.EXE on
those images? It always complained of not being able to find the boot file
when I tried using it.
Once you get logged in to VMS, I think it should be possible to use EXCHANGE to
poke around the console tape images on the file level. It might be possible to
find the reason for the problem that way?
I never got a login prompt, but perhaps that's because I booted with the
write protect switch on?
VMS would try to write to the pagefile at some point but I think it would
write a message (likely many messages!) on the console if that failed rather
than sit there doing nothing. It probably hasn't hit the pagefile yet.
It appeared to be trying to create or join a
VAXcluster for a while, then said something about loading MSCP disk server.
Before you do anything, make sure the ethernet network is properly connected
or terminated appropriately.
If configured to form a cluster, VMS will normally wait for a short period
for the other cluster member(s) to appear on the cluster interconnect
(usually the ethernet network) before continuing to boot. Try leaving it for
a few minutes and you may then get something like:
"VAXCluster state transition completed. Initialization continuing".
Or you might not. In a cluster, each node contributes a number of votes to
the cluster. None if the nodes will do anything until a cluster quorum is
present (more than half the number of votes usually present in the cluster).
If the machine just sits there indefinately after loading the MSCP disk server,
you probably don't have enough cluster votes to proceed and the best thing to
do is perform a conversational boot which usually involves setting the least
significant bit of register R5 to 1 before booting. How exactly to do this
varies from processor to processor and I don't know how to do it for an 11/730.
When you manage to do this and try booting again, you should get a
"SYSBOOT>"
prompt and you could:
SHOW VAXCLUSTER
SHOW VOTES
SHOW EXPECTED_VOTES
SET VAXCLUSTER 0
SET WRITESYSPARAMS 0
CONTINUE
to confirm that the issue is with cluster votes, turn off clustering and
proceed with the boot process.
I have plenty more experimentation ahead, including
seeing what's on that
RL02 pack labeled something like "VMS53SYS" (if I recall correctly).
My attempts to boot up the v5.3 standalone backup tape images I downloaded
still haven't succeeded. As suggested, I'll see if standalone backup might
be on another partition next time I work on the machine. I'd like to try
backing up both the R80 and the RL02 to tape if I can.
Not exactly a partition, more of use one of a number of different root
directories in the same partition to start searching for files from.
Eventually, I'd like to run an older version of VMS than 7.3 on it.
Preferably, something contemporary to when the 11/730 was still sold, or
at least from before any sane 730 users upgraded to newer and faster VAXen.
Of course, that assumes I can procure suitable installation media, or usable
images with which to create it.
I would suggest V5.5-2. This is likely newer than when the 11/730 was sold
but I suspect that many an 11/730 would have been upgraded to this version
while in service and ended their days on that version. It was regarded as a
good stable place to be with few unaddressed issues and it would be
increasingly difficult to get useful items such as a stable TCP/IP stack
running on versions earlier than V5.5-2.
Regards,
Peter Coghlan.
12 Jun
12 Jun
10:36 a.m.
On Jun 12, 2015, at 04:01, Peter Coghlan <cctalk at
beyondthepale.ie> wrote:
If the machine just sits there indefinately after loading the MSCP disk server,
you probably don't have enough cluster votes to proceed and the best thing to
do is perform a conversational boot which usually involves setting the least
significant bit of register R5 to 1 before booting. How exactly to do this
varies from processor to processor and I don't know how to do it for an 11/730.
When you manage to do this and try booting again, you should get a
"SYSBOOT>"
prompt and you could:
SHOW VAXCLUSTER
SHOW VOTES
SHOW EXPECTED_VOTES
SET VAXCLUSTER 0
SET WRITESYSPARAMS 0
CONTINUE
to confirm that the issue is with cluster votes, turn off clustering and
proceed with the boot process.
Thanks, that works! I also turned off write lock, which makes it happier.
Wow, that boot sure takes forever. What the heck is it *doing* for all of that time? :)
And at the moment, it's still booting, as I sip my morning coffee. Just started
printing like heck and beeping... Ah, it's printing all of the licenses that have
terminated. Maybe I should have lied about the date? Looks like the hostname is PIKE. Sure
glad my iPhone boots more quietly.
VMS use not authorized on this node. I sure hope it won't enforce that before I can
try a backup!
Finally! A login prompt! And no clue about the passwords. Uh, how can I shut this beast
down without a valid login? !?
Great. :)
--
Mark J. Blair, NF6X <nf6x at nf6x.net>
http://www.nf6x.net/
10:44 a.m.
Found this:
http://labs.hoffmanlabs.com/node/1078
Unfortunately, console halt isn't implemented on the 725/730, so I think a hard power
off is my only option.
--
Mark J. Blair, NF6X <nf6x at nf6x.net>
http://www.nf6x.net/
10:57 a.m.
Doesn't control-P on the console halt the machine on a VAX-11/730?
On 12 June 2015 at 15:44, Mark J. Blair <nf6x at nf6x.net> wrote:
Found this:
http://labs.hoffmanlabs.com/node/1078
Unfortunately, console halt isn't implemented on the 725/730, so I think a
hard power off is my only option.
--
Mark J. Blair, NF6X <nf6x at nf6x.net>
http://www.nf6x.net/
11:22 a.m.
On Jun 12, 2015, at 07:57, Antonio Carlini
<arcarlini at iee.org> wrote:
Doesn't control-P on the console halt the machine on a VAX-11/730?
It brings up the console prompt, but the (H)alt command just prints the PC rather than
triggering a halt on the 725/730.
Next time I work on the system (Tonight? Or maybe tomorrow... depends on how cool it is in
the evening, as that box pumps out enough BTUs that I need to open windows!) I will try a
password reset as described here:
http://labs.hoffmanlabs.com/node/204
Oh yeah, I forgot to comment on this earlier:
On Jun 12, 2015, at 04:01, Peter Coghlan <cctalk at
beyondthepale.ie> wrote:
Once you get logged in to VMS, I think it should be possible to use EXCHANGE to
poke around the console tape images on the file level. It might be possible to
find the reason for the problem that way?
I can poke around at the file level with my RT11 filesystem tool:
https://github.com/NF6X/pyRT11
In fact, the image I've used to boot the system isn't exactly the one that you
fixed for me, but rather a new one I created using its files and bootblocks, with several
scripts added to try booting different root directories. On the 730, I don't think I
can modify R5 in the boot command. Rather, I would need to either manually type in the
whole boot sequence manually, or create a new boot script off-system. I chose the latter,
and made a bunch of them to try probing different root dirs on both the R80 and RL02
drives. BTW, standalone backup was not helpfully installed on E on my R80, so once I
crowbar my way in, I'll try running a backup under the full OS.
And new replies have arrive while I was typing:
On Jun 12, 2015, at 08:00, Peter Coghlan <cctalk at
beyondthepale.ie> wrote:
You can get a free hobbyist license if you join whatever DECUS is called now.
I joined a while back, but haven't requested license yet since I didn't have my
system serial number handy. I'll pull the rack out and jot down the number before I
leave for work this morning.
[and an alternate set of password reset instructions]
Thanks! This is very helpful, as I am a real VMS noob. My previous experience with it was
just using it briefly in one or two classes, and running a canned script to perform
backups as a graveyard shift operator (Great pay for a student! And a key to the machine
room! And a staff account with no quotas, rather than buying limited time or fighting with
everybody else for CPU cycles on a class-account cesspit server! Do homework on a Sun
workstation instead of a Wyse terminal! Heck, run a sim on the C240 supercomputer!
Giggle!). I can still hear that TU77 howling in my mind. Hmm, I wouldn't mind having
one, with a matching 11/780... :)
Back in the day, I really hated VMS for no other reason than I liked UNIX and embraced it
with the natural snobbishness of a youth growing up in the computer environment of the
80s, where our computer was the "best" even though the other guy's
"crappy" computer used the same 6502 running at the same clock speed (but mine
really was the best, because it used a 6809 :) ). But now I want to learn about VMS and
appreciate it for what it is.
On Jun 12, 2015, at 08:11, Johnny Billquist <bqt at
update.uu.se> wrote:
Avoid V5, though. I remember at the time that people was having serious issues with that
version. V6 improved things again. I think DEC spent a fair amount of time to improve
performance because of all the complaints about V5.
V4 would also be good in some ways, but it's old and might feel limited if you want
some modern software running...
Modern software running on a 730... I don't know if I'll live long enough to wait
for it to launch! :)
--
Mark J. Blair, NF6X <nf6x at nf6x.net>
http://www.nf6x.net/
13 Jun
13 Jun
4:42 p.m.
I'm having trouble with the password reset procedure (but will resolve it by the end
of this message). When I run AUTHORIZE, I get this:
$ run authorize
run authorize
%DCL-W-ACTIMAGE, error activating image SECURESHRP
-CLI-E-IMGNAME, image file PIKE$DQA0:[SYS0.SYSCOMMON.][SYSLIB]SECURESHRP.EXE;2
-SYSTEM-F-PROTINSTALL, protected images must be installed
I tried installing my new OpenVMS license pak to see if that would influence it, but it
didn't seem to help. Took about an hour for my terminal emulator to slowly type in
that 2400 line license file!
Ok, let's try rebooting with the new license pak installed... Nope!
>>@DQ0GEN
>>I
>>D/P/L F26200 86
>>D/P/L F2620C FFFFFFFF
>>D/P/L F26200 6
>>D/G/L 0 00A80003
>>D/G/L 1 3
>>D/G/L 2 3FB86
>>D/G/L 3 0
>>D/G/L 4 0
>>D/G/L 5 1
>>E SP
G 0000000E 00000200
>>L/P/S:@ VMB.EXE
>>S @
SYSBOOT> SET VAXCLUSTER 0
SYSBOOT> SET /STARTUP OPA0:
SYSBOOT> SET WRITESYSPARAMS 0
SYSBOOT> CONTINUE
%SYSBOOT-I-SYSBOOT Mapping the SYSDUMP.DMP on the System Disk
%SYSBOOT-I-SYSBOOT SYSDUMP.DMP on System Disk successfully mapped
%SYSBOOT-I-SYSBOOT Mapping PAGEFILE.SYS on the System Disk
%SYSBOOT-I-SYSBOOT SAVEDUMP parameter not set to protect the PAGEFILE.SYS
OpenVMS (TM) VAX Version V7.3 Major version id = 1 Minor version id = 0
%WBM-I-WBMINFO Write Bitmap has successfully completed initialization.
$ set noon
set noon
$ spawn /nowait sys$system:startup.com
spawn /nowait sys$system:startup.com
%DCL-S-SPAWNED, process SYSTEM_1 spawned
$
%DCL-W-NOLBLS, label ignored - use only within command procedures
\SYS$SYSTEM:\
%DCL-W-PARMDEL, invalid parameter delimiter - check use of special characters
\.COM\
$ set default sys$system
set default sys$system
$ run authorize
run authorize
%DCL-W-ACTIMAGE, error activating image SECURESHRP
-CLI-E-IMGNAME, image file PIKE$DQA0:[SYS0.SYSCOMMON.][SYSLIB]SECURESHRP.EXE;2
-SYSTEM-F-PROTINSTALL, protected images must be installed
$
Ok, now let's try the slightly different procedure at
http://labs.hoffmanlabs.com/node/204 <http://labs.hoffmanlabs.com/node/204> with the
vaxcluster stuff spliced in... Much better! Much longer boot time, lots of complaining
about terminated licenses (previous attempt to install them must not have worked), but I
was able to reset the password. I'm not sure what was different about the other
proceduer (maybe the /nowait flag?), and some of the lines don't even look applicable
to a non-workstation.
Now running the license pak script again, and the output looks a lot more promising.
Forward progress continues!
--
Mark J. Blair, NF6X <nf6x at nf6x.net>
http://www.nf6x.net/
3308
days inactive
3309
days old
5 comments
3 participants
participants (3)
-
arcarlini＠iee.org -
cctalk＠beyondthepale.ie -
nf6x＠nf6x.net