disk i/o per process monitor

Hi,

I'm looking for a solaris equivalent (I'd assume maybe d-trace based) for
VMS's "monitor process/topdio".
Specifically we have some T3s with a bunch of zones that spend most of their
time thrashing the swap partition.
We're fairly sure that the culprit is a new version of a financial app running
in the zones but we need to identify which part. It has an embedded sybase
instance on each zone plus the app itself.

Even if we have a large amount of free memory the thing still kills the disk.

The goal is to find a way to identify which component in the setup is causing
the problem. The financial app has a large number of subsystems. None of these
stands out in terms of per-process CPU load. Older versions of the app using
the same version of embedded sybase on the same hardware don't exhibit the
problem.

Any suggestions gratefuly accepted.

Apart from "go buy some VMS". :-)

Cheers

Joe