On Thu, Dec 02, 2004 at 02:23:33PM -0500, Brian Cuttler wrote:
> What is MFM ?
Drive technology from the late 70s and early 80s.
Strictly, MFM is "modified frequency modulation", one of several
algorithms for converting between a digital bit stream and an
analog signal that you can actually record on magnetic media (a
couple of others are (unmodified) FM, RLL, and GCR). MFM was
used on a lot of hard drives (aka "Winchester disks") back then,
and is still used on floppies. Newer hard drives have more
efficient algorithms to do the same task.
The Seagate ST-506 was a famous example of a drive that used MFM
recording -- 5 whole megabytes in a 5.25" full-height form
factor, which seemed pretty darned impressive at the time. (BTW,
the 5.25" drive bays in a modern PC case are half-height (though
some people erroneously call them "full-height"); a true
full-height drive would occupy *two* bays.)
Less formally, "MFM" refers to the interface that was used by the
ST-506. That interface was copied by many other drives, to the
point that it became a de-facto standard; in this sense, it's a
peer of IDE and SCSI (though the interface itself was much much
*much* more primitive than either of those).
(MFM drives happen to be on my mind these days, because I'm
helping a friend to try to revive an PDP-11-based system from
back then, and rescue the data off of its decrepit ST-506's onto
something less likely to die at any moment :-/)
> On a more serious note. Ran amdump on this server yesterday and the
> amstatus only showed 13 partitions, I know there are 14. At least that
> is all that showed while I was watching the estimation phase. I was able
> to see the 14th DLE listed in the log files though and all 14 where
> listed in the amdump report when I received it.
I wouldn't worry too much about this. Amstatus does its thing by
reading tea leaves (specifically, by poring over the
$logdir/amdump output file), not by querying the running Amanda
processes. It's hardly authoritative.
Maybe you've found a buglet. If it happens again, grab the
"amdump" file that causes it (quickly, before more records are
written). That'll help someone to figure out what's going on.
> I'm a little surprised to see that we have no concurrency, even when the
> first few relatively small partitions partitions where being dumped.
See "maxdumps" and "inparallel" in amanda.conf, and "spindle" in
disklist.
Also, amstatus will tell you, down near the bottom of its output,
why no more dumps can be started at the moment.
--
| | /\
|-_|/ > Eric Siegerman, Toronto, Ont. [EMAIL PROTECTED]
| | /
The animal that coils in a circle is the serpent; that's why so
many cults and myths of the serpent exist, because it's hard to
represent the return of the sun by the coiling of a hippopotamus.
- Umberto Eco, "Foucault's Pendulum"
