A wild guess but worth mentioning...

Is the patch MPELXW5B installed? This patch supersedes some critical patches
for large memory support on 7.0. I know of two cases where this patch failed
to qualify during patch installation however the symptoms of its absence were far
more severe than what you describe.

The patch MPELXW5B is on 7.0 power patch 1 and itself addresses a problem
with TC (transfer of control) on N-class multiprocessor systems. This patch super-
sedes (and therefore includes all of) MPELXR7 and MPELXW0. These two
patches address problems found in I/O virtual addressing. IOVA is used only when
memory is > 3.75GB. There are a few other patches that are also superseded by
MPELXW5B but these IOVA patches would be required on systems with more
than 3.75GB.

HP support is presently investigating why MPELXW5 failed to install in the 2 cases
we have seen so I don't have any details on that right now.

HTH,

Bill
HP/CSY
reply to: cadier@[hpixec01].cup.hp.com

----- Original Message -----
From: "Cecile Chi" <[log in to unmask]>
To: <[log in to unmask]>
Sent: Monday, November 19, 2001 3:58 PM
Subject: [HP3000-L] 7.0 Performance hit?


> I was expecting to see a performance hit with 7.0, but our MRP runs
> were worse than I expected.
>
> HP3000/960, running MPE/iX 6.0: 180 minutes
> (running alone)
>
> HP3000/969(3.75 Gb memory), running MPE/iX 6.0: 51 minutes
> (other jobs running)
>
> HP3000/N220 (4.5 Gb memory), running MPE/iX 7.0: 155 minutes
> (running alone)
>
> The N-class box is supposed to replace the 960; I guess it will do that.
> It will be interesting to see what happens to MRP when the 969 is
> upgraded to 7.0
>
> Cecile Chi
>
> * To join/leave the list, search archives, change list settings, *
> * etc., please visit http://raven.utc.edu/archives/hp3000-l.html *
>

* To join/leave the list, search archives, change list settings, *
* etc., please visit http://raven.utc.edu/archives/hp3000-l.html *