• Please review our updated Terms and Rules here

Dumping Images of my VAX-11/730's Drives?

NF6X

NF6X

Veteran Member
Joined
Sep 9, 2013
Messages
1,559
Location
Riverside, CA, USA
It's supposed to be cold and rainy tomorrow, so I'm thinking that it might be a good time to fire up my VAX-11/730 again. If I remember correctly, it has OpenVMS 7.3 on the R80, and VMS 5.2 on an RL02 pack. After my previous adventures getting it to bootstrap from tu58em running on a Mac, I can boot the system from either drive.

I want to find some way to dump out images of the drives and get the data onto my modern machines before changing much of anything, other than the changes I've made already to update VMS licenses and so forth. Once I manage to image the current contents of the drives, I'd like to start from scratch installing some VMS version that's roughly contemporary to the machine, and then build up from there as I learn VMS.

I've made tape backups of unknown quality on the system's TU80 drive, but I do not yet have any way to read those tapes on my modern machines. I've also spent one afternoon trying to bring up TCP/IP networking several months ago, but I didn't succeed. I don't know yet whether the ethernet card, AUI cable and 10base-T transceiver are good. I don't know for sure if the right software is on the machine to talk TCP/IP, though it did seem to be responding in a sane manner to my attempts to bring up TCP/IP. And on top of that, I'm a real VMS noob, so I still have a big learning curve to climb. All I know so far is that I didn't manage to get a ping to work in either direction.

I have a broken 9-track drive with a SCSI interface that I plan to fix one of these days, so maybe that will provide a way to get data off the VAX and onto my modern machines. But I think it'll be a while before I fix the drive and get it talking to one of my "newer" machines with SCSI interfaces (Sun Ultra 60 running Solaris 8, Amiga 3000, or a ratty old 386 clone running DOS 6.22).

I have a TK50 drive and a SCSI TK30 drive; I was thinking of rigging up the TK50 to the VAX and hooking the TK30 to one of my newer machines. But trying to get the TK30 to work has been a dismal failure so far. I can share pictures of oxide-shedding tape tangled up inside the drive if anybody would like.

I wonder if any of the VAX/VMS experts here might have suggestions about good ways to go about dumping the contents of these disks? Should I continue trying to make TCP/IP work, or would I be better off trying to bring up DECNET and setting up some other machine (maybe a Linux box?) to act as a bridge between the VAX and my modern systems? Or maybe I should try yet some other scheme to hook up some other form of portable storage to the VAX? Hmm, I don't recall whether I acquired a UNIBUS SCSI card before I closed down my eBay account last year. If I have one in the pile, maybe that will offer options, such as hanging a SCSI2SD off the machine.

Assuming I manage to interface portable storage to the VAX or get a suitable communication channel working, I'll probably also need some help learning how to dump raw blocks from the drives, e.g. doing something equivalent to "dd if=/dev/somedisk of=somefile". I really don't know my way around VMS yet, and I find it a whole lot less intuitive than UNIXes. But I do want to learn VMS, and that's one of my reasons for playing with this cool VAX!

So, can anybody suggest a good direction for me before my next bout of exploratory fumbling?
 
Hi All;

NF6X, I know nothing about VAX/VMS ..But my self, I would and have gone the 9 track tape route.. It's been some time ago, but, I was able to read and write on my 9 Track tape drive from my PC.. Using SCSI card in my PC.. I used to be able to get old Cards from my local Electronic's store, but they went unexpectedly belly up.. My tape drive has the option of being SCSI as well and using some Linux MT commands made it work fine.. So, I would suggest getting Your Tape Drive(s) to work..
"" I can share pictures of oxide-shedding tape tangled up inside the drive if anybody would like. ""
If these are Older or tapes that have sat for a long time.. You need to Bake them before using them, OR the Oxide Will come off.. I know I have had a tape of Clear tape, that is minus the oxide, from trying to save a Non-baked tape.. I have forgotten what the Temperature needs to be but, I would say about 125 degrees, but check with others who Know for sure before trying to bake You Tape(s) No Whip Cream topping needed.. I have asked someone who knows and I am waiting for an answer..

THANK YOU Marty
 
Last edited:
Finding a Unibus SCSI card that emulates MSCP would give the most flexibility. The Emulex UC17 would be my choice if you have a open Unibus slot and the funding. You an create bootable images using an SCSI2SD and use 'dd' to save the microSD cards and mange them via another *nix like machine. You would also get a decent media sizes compared to the RM80 or RL02.

The other options don't have as much flexibility. You should be able to cluster the 11/730 with a MicroVAX 3100 or similar and use its SCSI interface for backups, if you use the VMS 5.2. But if you lose the VMS 5.2 boot disk on the 11/730 itself, I don't believe you can netboot the 11/730. Also clustering under 11/730 may not work under VMS 7.3

You do have to be careful with the different bad block mapping techniques on DEC media or emulations. Using 'dd' comes with it own gotcha's as this article points out - see section Copy a drive to an image file in http://www.noah.org/wiki/Dd_-_Destroyer_of_Disks


I've lost confidence in tape media these days. I keep my master copies on both rotating magnetic and ssd type media, and rotate to newer media regularly.
 
Fixing the SCSI 9-track would also provide a potential way to get new stuff onto the system, of course. I'll probably need to do that in order to make media to allow me to install a new OS on the VAX, anyway. I have two Kennedy 9610 drives; one with a Pertec interface, and one with a SCSI adapter card. Both have hub motor drive problems, and will require component-level debug. It is my eventual goal to fix one or both and find some way to interface it to some machine that gives me a path to/from my modern machines.

I would certainly like to have a UNIBUS SCSI adapter. I never managed to acquire one before parting ways with eBay, and I wouldn't be willing to pay a grand for one. I see one UC17 on eBay at the moment for $949 OBO, but that's more than I'm willing to spend. I did manage to find a QBUS SCSI adapter for use with my PDP-11/03 machine, but I have had trouble getting it to work. I seem to recall from my last session with it, several months ago, determining that I'd probably need to swap in a PDP-11/23 CPU for compatibility with it.

If I had a way to read and write RL02 packs on a machine with a path to my modern machines, then that would also give me interesting options. I have a pallet of RL02 drives, with plans to eventually use one or more of them in my PDP-11/44 project. That's a long-term project, though. I think I'm more likely to use the VAX to dump and write RL02 packs before I get any other way to do that, because it's the "workingest" of my old DEC machines already. For that matter, the VAX might end up being my system for reading and writing 9 track tapes, too, if I get the networking up and running.

Clustering with a newer MicroVAX sounds like an interesting approach, and might be a cheaper option than finding a UNIBUS SCSI card. I'd need to find one in close to turn-key condition, of course, or bringing it up could be a bigger project than the VAX data transfer issue that it's supposed to address. I'll keep this in mind in case I happen to stumble over a suitable MicroVAX system.

I think that bringing up networking on my VAX is the only approach that I would have any chance of accomplishing today. But even if I bring up networking, whether TCP/IP or DECNET, I do not yet know how to use it to get raw disk images off of the machine under VMS.
 
If you already have a QBUS SCSI controller + 11/23, then adding a SCSI2SD and a RLV12 would allow you to make copies of your RL02 Packs.
 
Hmm, that's right. Naturally, I don't think I have an RLV12; I just have an RL11. I can add that to the be on the lookout list.

Yet another option would be some sort of Pertec adapter for one of my machines that have a path to my modern machines. That could allow me to borrow the TU80 drive from my VAX. I'm not positive, but I think that the TU80 drive has a Pertec interface.
 
I have some Chi / Computer Logics PCTD-16 ISA bus Pertec controllers, but you don't do PCs and DOS. :)
 
gslick said:
I have some Chi / Computer Logics PCTD-16 ISA bus Pertec controllers, but you don't do PCs and DOS. :)
Click to expand...

Is there good software to run them under DOS? I have an AT-bus 386 clone that I use for running ImageDisk.

I just dug out an old Celeron 733 system that I luckily never got around to throwing away. I may be able to install a suitable older Linux version on it with DECNET support, then use it to log into the VAX. Once I do that, maybe I can just log hexdumps of volumes using DUMP on the VAX, and then write some hacky Python program to reconstitute the dumps into image files? I'm not sure how I would transfer stuff in the opposite direction, but it would be a start. I think it's a PCI bus machine, so I could potentially use a PCI Pertec card in it, too.
 
I've never seen a PCI bus Pertec interface card.

The PCTD-16 is a 16-bit ISA bus card. I have never really used the application that came with the card. It also has a TSR driver with a documented interface for reading and writing blocks. I wrote simple applications to use the TSR driver to read a physical tape into SIMH style .TAP images, and write SIMH style .TAP images back out to physical tape.
 
Hi All;

Gslick, I might have a card that is similar to Yours, I can't get to it today to see who made mine, but, I should be able to tomorrow.. The problem I had with mine was, that If things were not just right, speed of the Computer, right or correct port assignments there were some choices, just not enough, I could not get it to Work,
and so I went to a ScSi controller.. The Drive has a Pertec output, and I have a Pertec to SCSI adapter made specifically for this tape drive..
Anyway, what can You tell me about Your Pertec Adapter ISA card ??

THANK YOU Marty
 
I've made some progress on the DECNET front. DECNET support is still in the package repositories for Ubuntu Linux 14.04, so I set up a VM on my Mac to try running it.

The bad news is that I can't get the VAX and the Linux VM talking to each other in a meaningful way. It might just be PEBCAK, but I'm getting the impression that the DECNET applications on the Linux end may have suffered bit rot since their support was abandoned.

But the good news is that the Linux VM and the VAX detect each other's presence on the network, so at least it seems that the VAX's ethernet hardware is working. So once I work out something better on the modern computer end, I should be able to get the bytes a-flowin'.

The OpenVMS 7.3 installation on the R80 fixed drive looks like it was just put on for hobbyist fun, but the VMS 5.3 installation on an RL02 pack looks like it came out of a working environment. The system was part of a pretty large DECNET network with a few dozen nodes in the known nodes table. There are COBOL and FORTRAN licenses on the system, but I don't know if the compilers are present, or maybe just the runtimes. I don't really know what to look for yet.

While I had the VAX running under each OS installation over the last couple of nights, I did recursive directory listings and printed out as much interesting info as I knew how to, so I have some nice thick stacks of fanfold from the DECwriter III for future reference.
 
One issue might be that DECnet Phase IV on ethernet requires you to set the MAC address on the physical ethernet interface based on the DECnet node number. This is because DECnet Phase IV does not use an equivalent of the ARP protocol used by TCP/IP to map IPv4 addresses to MAC addresses. I would expect this to cause problems for a virtual machine.

DECnet Phase IV address converter. For example, DECnet node 1.1 needs to have MAC address AA-00-04-00-01-04.

Part of Cisco page Troubleshooting DECnet is Cisco-specific, but a lot of it is basic information.
 
That's not the issue in this case. I'm using ethernet bridging in VMware Fusion, and the MAC address changes just fine. The Linux VM and the VAX both see each other as reachable nodes, but I'm having trouble higher in the stack.

Now I'm trying running an older Linux distribution in a VM, from a time when the DECNET code was still being supported. Still monkeying around with kernel builds trying to get it to even see the virtual network adapter...
 
IIRC, DECnet nodes finding each other on ethernet is done with a multicast address (not the actual node MAC). So I don't think an adjacency actually proves that the Linux box can receive packets addressed to the DECnet node MAC. But that certainly doesn't mean that is the problem, either.

Can you get anywhere with NCP LOOP?
 
Success!

Running Debian "Woody" in a VM under VMware Fusion on my Mac Pro. I had to mess around with building different linux kernel versions for a while to get networking to see a network adapter at all under the VM. I knew I was onto something when, after the latest attempt to build a kernel and modprobe the right driver, I typed "ifup eth0" on the VM and head the DECwriter III in the other room beep and print out a message about the VAX seeing a new node appear. After that, it only took a few tries to get the VAX reading and writing files on the Linux VM!

It's well past time for bed, but I think I made a big breakthrough tonight.

Cda7jZfUYAA994x.jpg-large.jpg Cda8gL_UAAAwces.jpg-large.jpg
 
Hi All;

THAT's GREAT !!!!!!!!!!!!!! A Great Job, well Done..
Now can You Explain in Plain English what You did, so others can do it as well.. And, so it is Documented, and laid down for others to do..

THANK YOU Marty
 
You must log in or register to reply here.
Back
Top