cctech Digest, Vol 29, Issue 1
mark at matlockfamily.com
Tue Nov 1 21:20:45 CDT 2016
> Date: Mon, 31 Oct 2016 10:48:56 -0400
> From: Paul Koning <paulkoning at comcast.net>
> To: "General Discussion: On-Topic and Off-Topic Posts"
> <cctalk at classiccmp.org>
> Subject: Re: For sale 128K Core RAM Industrial PDP 11/40 Massbus
> Message-ID: <A0F4A035-0EAA-4343-87DC-86495173458F at comcast.net>
> Content-Type: text/plain; charset=us-ascii
>> On Oct 31, 2016, at 10:26 AM, william degnan <billdegnan at gmail.com> wrote:
>> Given 128K core, wouldn't one be able to save the OS in core, no need to
>> load what would need to "get started" from a diskpak? The data would be on
>> the tape drives, and something on stand by to re-load the OS back into core.
> Only if the OS implements the ability to resume from a power failure without reloading from disk or tape. Few do. Perhaps some flavors of RSX, I don't know. RSTS V4A, when built with the power fail handling option, could do so. Later versions do not; they unconditionally reboot (from disk) at powerup.
RSX-11M and M+ did resume from power failures very robustly. I had a PDP-11/44 with batteries for the MOS RAM (and fans to cool it) with RK07 drives. I can remember testing it by pulling the power with a number of applications runnings including editors etc. The disks would yank the heads back and spin down, then when the power was restored and the drives spun up, it was like it never happened. I might see an entry or two in the error log file about a disk retry, and a user might loose a keystroke on the file they were editing but it absolutely was solid. I don't know of a single modern operating system that can do that today.
RSX-11M+ can run TCP/IP today with Johnny Billquist's BQTCP package but it dies require I/D space processors. M+ itself can run on a 11/23+ or 11/24 but it really needs more than 128KW and the lack of I/D space really limits its capabilities.
More information about the cctalk