Tnaks to all.
Gary
----- Original Message -----
From: "Tom Renz" <[log in to unmask]>
To: <[log in to unmask]>
Sent: Friday, January 05, 2001 10:47 PM
Subject: [HP3000-L] FWD: Re: Re: [HP3000-L] Transact database program
getting slower


> Oops, I forgot to mention that the f-r-e-e utils are available from
TRenzSoft.  Don't know if there are similar ones available from the CSL.
The program names I have are "MSTRMAP" (the Primary vs. Secondary graph),
"DBINFO" (a lot of combined info regarding a data set), and "IMAGEVER" (set
variables for the TurboIMAGE current version and the individual db calls).
>
> Let me know if you are still interested.
>
> Tom
>
> ---------- Original Message ----------------------------------
> From: "Art Bahrs" <[log in to unmask]>
> Reply-To: "Art Bahrs" <[log in to unmask]>
> Date: Fri, 5 Jan 2001 07:52:16 -0800
>
> >Hi Tom :)
> >   Ok... so a bit of a <PLUG> for the Interex CSL... are the free utils
in
> >there? if so what are the names so I can give 'em a try too :)
> >
> >Art "Free is a very good Price! (tm Tom Peterson) Portland Icon hehe"
Bahrs
> >
> >----- Original Message -----
> >From: "Tom Renz" <[log in to unmask]>
> >To: <[log in to unmask]>
> >Sent: Friday, January 05, 2001 7:06 AM
> >Subject: Re: [HP3000-L] Transact database program getting slower
> >
> >
> >> Chris,
> >>
> >> We have a no cost (f r e e) utility that will map any master data set
> >> showing a graph of the Primaries and Secodaries and the percentages of
> >> each within each percent of the set's capacity.  We will be happy to
> >> send you a copy to see what your distribution and where your peaks are
> >> within the data set(s) in question.  Let us know.
> >>
> >> Also we have a couple of other no cost utilities that set vars with the
> >current version of TurboIMAGE and if desired each db intrinsic.
> >> Basically the same list as listed in QUERY but setting variables.  And
> >> the second no cost utility lists all data sets within a data set
> >> and almost all the info you would like to see for each data set
> >> (name, set number, MPE file name, Extent names if jumbo, the number of
> >> sectors used, the record length, media length, the data sets tied via
> >> Image Keys, average chain lengths, and more).  Let us know if you are
> >> also interested in any of these other no cost (f r e e) utilities.
> >>
> >> <Plug>
> >>
> >> Now for the plug.  We also have a tool that can measure each db call
> >> made and provide statistics on the calls to provide you information
> >> such as longest call in CPU & wall time.  This could answer how long
> >> it takes for each DBPUT to complete as well as any other db call.
> >> Let us know if you would be interested in a demo.
> >>
> >> </Plug>
> >>
> >> Tom Renz
> >>
> >> ---------------------
> >> On Wed, 3 Jan 2001 at 05:46:49 you wrote:
> >> >Hello everyone
> >> >
> >> >Thanks for the replies.
> >> >
> >> >My master set is 3 million records in a capacity of 6 million.  The
> >> >set is therefore only 50 percent full.
> >> >
> >> >The key item (SYGEKSNR) X10 is a 10 digit number built as follows
> >> >DDMMYYNNNN.  The whole thing satisfies some modulo 11 arithmetic.  It
> >> >could be possible that I am getting secondaries.  How can I check ?
> >> >Are there tools available for checking ?
> >> >
> >> >I am splitting my program into updating and adding and will perform
> >> >some timing tests.  I really don't know where the bottleneck is.
> >> >
> >> >The updates to the database are received every 15 minutes.  How safe
> >> >would output deferred mode be in this case ?
> >> >
> >> >I am using entry level locking to update existing records.
> >> >
> >> >We have 37 machines with varying memory configurations and Operating
> >> >system versions, but the majority are 6.0 or higher.
> >> >
> >> >Thanks again for the help
> >> >
> >> >Chris
> >
> >