Re: Creation of 10000's of empty table segments and more...

> Also, there are no core files lying around anywhere, which is somewhat
> surprising.

Hm.  On some platforms, processes started from system startup scripts
run with "ulimit coredumpsize" set to 0, which prevents coredumps.
You may need to start the postmaster manually with a normal ulimit
before you will get a corefile.

> Does it make any difference that I compiled with a BLCKSZ of 32768 and
> NAMEDATALEN of 64?

In theory, no ...

> All I do have is the intact contents of the database directory of the
> problem database.  Is there any way to move this to another
> installation so that I can have a look at it, and maybe get you a core
> dump or at least a detailed log on another machine?

If you have the entire $PGDATA directory, you can copy it to another
identically-configured machine.  Or you can leave it where it is and
start up a test postmaster in it (just specify -D pointing at the data
dir, and select a port number other than the standard 5432 with -p).
If you do that, just remember to specify the special port number when
connecting with psql (use -p, or set environment variable PGPORT).

A coredump backtrace should help, especially if you recompile with -g
first.

> Also, would upgrading to 7.0.2 help,

Possibly, but it's hard to tell until we know more.  I'd suggest leaving
your installation alone until we have more info.

            regards, tom lane