27 Jan
2018
27 Jan
'18
8:14 p.m.
On Jan 27, 2018, at 11:01, Rob Jarratt
<robert.jarratt at ntlworld.com> wrote:
I run SIMH on a couple of Raspberry Pis without issue, one running VMS 5.4 the other VMS
7.3. They each run as cluster members so I have a cluster for 5.4 and 7.3 running at all
times. It is very straightforward. I have even set it up to automatically boot SIMH when
the machine is powered on using a tool called ?screen?, so I can run it completely
headless. I have found that the Raspberry Pi networking starts to get a bit unreliable
after running these machines 24 hours a day for a few years.
I have SIMH running on a BeagleBone Green, though I don't have a lot of time or use on
it yet. It's just the native DECnet support under Linux which doesn't work out of
the box using the source in the kernel distribution and the userland stuff available via
apt-get.
One thing I'd like to look into for a full-time server is whether I might be able to
have host system shutdown trigger a VMS shutdown on the emulated system, or alternately,
just halt and checkpoint the emulation. That could be nice in conjunction with a
supercap-based backup supply for the host computer, so it can have time to shut down when
the power unexpectedly drops. I have an MFM hard drive emulator designed by David Gesswein
which is built around a BeagleBone, and it has supercap-based backup to let it be embedded
inside an old computer and simply powered on and off without worrying about performing a
shutdown on it first. When the power gets yanked, the supercaps keep it running long
enough to shut down and dismount its filesystems cleanly.
Is the network unreliability that you mention just a matter of high uptime? Does an
occasional reboot fix it, e.g. by resetting memory leaks?
--
Mark J. Blair, NF6X <nf6x at nf6x.net>
http://www.nf6x.net/