[rescue] uVAX II funtimes

[John Francini]

On 27 Jul 2009, at 14:01, Ian Finder wrote:

>
> The only qbus card other than the CPU is the DEQNA ethernet card,
> which, to my understanding, is the less preferable of the two cards
> commonly found in such a unit.

Yes, but given the overall speeds involved, it'll be good enough.

How much memory is there?

>
>
> Still, it's probably a very good thing I have it as there is no hard
> disk, the floppy drive was disconnected and is missing its cable, and
> the disk controller was given to me separate from the chassis and is
> in unknown condition. I should also make clear I don't have a hard
> drive, and I have no idea what sort of drive I'd even need, though I
> haven't researched the topic thoroughly.

The disks you'd want to find are RD53 (73 mb) or RD54 (155 mb) MFM  
drives. RD52 drives (30 mb) will work, but are tiny even for that time.

There are two types of disk controllers for internal drives: the RQDX1  
and the RQDX3.  You want the RQDX3; the RQDX1 was ghastly slow even in  
its target application -- the MicroVAX I -- which ran at one-third the  
speed of the MicroVAX II.

>
> For now, instead of tackling a pile of 25 year old hard discs, I was
> thinking I could netboot the MicroVax and operate it completely
> diskless. I *believe* this is easy with NetBSD, but I really want to
> try VMS. Is this a retarded idea? Will it even work, and if it does
> will it totally sack my performance?
>

Well, the biggest problem you'll likely have is dealing with the boot  
protocol.  Since DEC used DECnet, it had its own boot protocol, MOP,  
which requires a MOP server on any host that's going to boot it, along  
with the multiple files that are needed to boot it.  The firmware does  
*not* know how to speak bootp.


> It's my understanding that VMSclusters have provisions for diskless
> clients beyond installation and maintenance, so I was going to attempt
> this by setting up a PC networked with the Vax (the decnet port is
> just standard AUI, correct?) and running VMS under SIMh. The diskless
> Vax would then be able to run VMS as well as any standard, uncrippled
> machine, right? Or am I way off-base and describing a
> performance-tanking deadweight for a machine with such limited
> resources?

Yes, it's standard AUI.

I've used diskless VAXstation 2000 systems (based on the MicroVAX II  
CPU technology) in a cluster, back around 1989. The only way to make a  
diskless VAXstation perform reasonably well was to put a small local  
hard drive into the thing for use as a paging/swapping disk.  And even  
that was seriously compromised over the 10-megabit networking of the  
day compared to a locally-booted system.

If your PC is fairly recent, it will likely run VMS significantly  
faster--in emulation--than the MicroVAX II will.

Just my $0.02,

John
(who ran a MicroVAX II/VAXstation 2000 cluster in a former life at DEC)