1.  TPI is turned off.

2.  There are no chains since these are stand alone detail datasets.

Carl

-----Original Message-----
From: Don Cotton [mailto:[log in to unmask]]
Sent: Monday, November 01, 1999 1:59 PM
To: Carl McNamee
Subject: Re: Performance question


Carl,

A couple of things to look at :

1.  Are you using any kind of TPI?  If yes turn it off and then reindex
after the databases are loaded.

2.  Are there any sorted chains?  If yes try sorting the flat files so that
you can minimize the amount of time Image spends on chain mgmt.

Don

At 01:22 PM 11/1/99 , you wrote:
>We ran into a unique situation this weekend while testing some batch jobs
>that will be used to perform a database conversion next weekend and I would
>like to see what the gurus of the list think could be causing us grief.
>
>We have 24 databases split between 6 volume sets.  We are extracting the
>data from then into flat files that reside on 8 different volume sets.  So
>far the process worked very smoothly.
>
>Next we put all the records from the flat files back into databases.  What
>we see on the system is that when running more that one of these batch jobs
>they start impeding on some resource.  We have eliminated the Put/Delete
>semaphore by having each batch job write to a different database.  Disk and
>memory utilization is very low so I do not believe that we have a bottle
>neck there.  Global CPU is only at 40% during this time.
>
>My thoughts are that we are having problems with a global semaphore but I
>cannot figure out which it may be or if there is a possibility of working
>around the problem short of running 1 or 2 batch jobs at a time.  This is
>happening on both a 996/800 and a 969/400.  Both systems are on MPE 6.0.
>
>Any thoughts on this issue will be appreciated.
>
>Carl McNamee
>Systems Administrator
>Billing Concepts
>(210) 949-7282

Don Cotton
Output Development Manager
ROC Software
512.249.9294    512.258.6968 (Fax)