[cap-talk] Styles of persistence

[Jed Donnelley]

At 07:06 PM 4/3/2008, Bill Frantz wrote:
>jed at nersc.gov (Jed Donnelley) on Thursday, April 3, 2008 wrote:
>
> >Perhaps Bill can explain what makes the KeyKOS space bank unusual
> >in this regard.  I'm afraid I don't know what a "space bank" is
> >or how it is organized internally.  Is that relevant to this
> >discussion?
>
>Quite. A space bank[1] is the object that allocates and rescinds
>all space in the system. (Think of it as the disk allocator for the
>file system.) Since the system has orthogonal persistence, it is
>running all the time, from when the system is first initialized to
>when, hopefully many years later, it is abandoned. It runs through
>checkpoints and restarts like all programs running in the system.

But just like any process (program?) on the system can't it be
restarted by some other controller?  Restarted with different
code that, as Jonathan noted, reads in it's old format data
and writes out new format data.  Surely the code for the space
bank changes over time.  If not then it seems to me you've
created a seriously unworkable system.  If so then it seems to
me you have an opportunity to do the "serializing" that
Jonathan suggests.

>Since it is responsible for space allocation on a single system, I
>don't think it makes much sense to serialize its state and move it
>to another system.

Not serialize to move in space but in time.

--Jed  http://www.webstart.com/jed-signature.html