Notes - database meeting - Feb 28, 2005

Present: Valeri Lebedev, Dennis Box, Lee Leuking, Mark Fischler,
         Norm Gelfand, Jean Slaughter

Purpose: 1) Review BLASTMAN status
         2) Define the future of the joint CD/AD project.

Next meeting: 1:30 Tuesday March 15, room to be determined.

Action items:  The goal is drafts of the 1st 3 items by the next
               meeting.)

         1. Write a note specifying the migration path to move
            the database from Dennis' desktop to a machine in the
            Integration Department and to turn over the management
            to Michele.  This should include a timetable. It also
            should include an estimate of the computing load and
            other requirements for disk and backup, so that the
            integration department can decide if it needs to add a
            machine.  Dennis and Michele will make the migration plan.
            Norm will estimate the load, etc.

         2. Write a note specifying the amount of Dennis' time needed
            to complete entering the remaining machines and transfer
            lines into the database.  (5-10 weeks over the next two
            years as the information is made available by AMG for the
            survey data and the AD liaisons with TD for magnet data.
            Dave Augustine's location data is the easy part.)
            - Norm, Dennis.

         3. Write a note specifying the access methods and reports
            that AD would like CD(Dennis) to provide. Norm, Dennis.

         4. Investigate moral or otherwise support for AMG.
             Jean, Valeri.


Details:

         1. Dennis reported on his progress and made some excellent
            summaries  - see the message appended below.

         2. Norm reported that there is a need for general purpose
            access methods for people who don't know SQEL.  He
            described the general functionality he thought was
            needed based on talks with Jerry Anala and other
            Tevatron users. Several possibilities ( and their
            problems) were discussed, including Microsoft ACCESS
            and MSWEB(?).

            We also discussed the usefulness of a tool to plot/list
            the location information for easy checking.  Ultimately
            there will be a need to check the lattice repository
            information against the database information.  Issues
            with respect to naming conventions were also mentioned.

         3. Norm also said there is a general problem in getting timely
            access to current survey data.

         3. Jean harped on the need to involve people like Elvin Harms
            (for pbar) and Eric Previs for the proton source, and ??? for
            transfer lines, both for entering data and for access tools
            and reports.

Dennis' message.

The current BLAST II database schema is documented at
http://odets3.fnal.gov/~dbox/p2_beams.html.
It can be browsed starting from
http://odets3.fnal.gov/~dbox/php/PSQL_PHASE2/browser/blast.php
An ER diagram, showing dependencies is at
http://odets3.fnal.gov/~dbox/p2_beams.ps

A few notes on the database contents:

The database currently contains location history,
harmonics, transfer constants  and quench
measurements for Tevatron and Main Injector magnets and spares.

All tables have naming convention of the form
't_some_descriptive_name', all columns names
are of the form  'c_some_descriptive_name'.
This is primarily done to avoid collisions with
reserved words from different database vendors but it
helps with general clarity in the SQL
and other access code.

The primary key to access magnet data  is consistently
c_magnet_name, and its contents should
be  generally recognizable to anyone who works with magnets
at Fermilab ' TB0211', 'TQ150D'etc.
The primary key to access locations is c_magnet_location.
There is a well agreed upon naming
convention for Tevatron locations, i.e. 'A11-1' .  I have less
confidence that the Main
Injector locations are named so that everyone agrees, i.e.
the same location is called either
H120 or MI120-2 depending on the data source.

Harmonic data, transfer constants,  quench current measurements
 and other multi-dimensional
measurements (i.e. we measure the first 14 harmonics at 6 currents
for TQ magnets but 5
harmonics at 2 currents for IDA magnets) are supported by this
schema.    Magnet data of this
type  from other machines such as the recycler should fit in with
no problem.    This schema
does not yet handle other desirable data elements like aperture
measurements, By(x),
drift/snap back,  etc,  but  can be extended to  handle them.