[Info-vax] Problem with RWMPB

[jbriggs444]

On Oct 8, 9:36 am, koeh... at eisner.nospam.encompasserve.org (Bob
Koehler) wrote:
> In article <4acdac39$... at glkas0286.greenlnk.net>, "Ian Dean" <ian.d.d... at baesystems.com> writes:
> > Hi,
> >     Whilst running an image, it hangs and shows a status of RWMPB (Modified
> > page writer busy).
>
> >     Can someone explain (in simplish terms) what is causing this and what
> > actions I can take to overcome this?
>
>    Your process needs to complete a page fault to bring a page from disk
>    into RAM.  In order to do that memory management is trying to make
>    space in your process' working set by moving pages out.  This has
>    caused the system wide cache of such pages to overfill so the memory
>    manager is trying to write some out to the disk, most likely to the
>    page file, but that isn't happening.
>
>    So most likely your page file is too small.  See if you can get in
>    a "show memory/files" from some other process.

If the system as a whole has ground to a near-stop with various
processes sitting in RWMPB and COMO states then yes, this would be a
classic "page file hang".

Most of the page file hang's I've seen turned out to be runaway
processes with memory leaks.  The challenges in such a case are being
able to get control of the system long enough find the offending
process and swing a big hammer (e.g. $ STOP /ID) and getting enough
resources free so that the target process can take the hint and die.