Subject: erd review
here ' s a summary of our conversation so you have points to follow up with .
general status of the database model :
fix not represented - need to incorporate into common design and standardize
- might cause some fix rework .
referential integrity not entirely enforced - potential for bad data to
develop .
snapshots from global databases are currently daily - this can be easily
changed to be more frequent , but need to consider implementation as
decision on direction of global as a part of commodity logic is made .
rate information not developed - need integration with rate server or mkm ,
preferably mkm . need to define how mkm will be used - whether just for index
names or to obtain actual settlement prices or curves as well .
application development is not currently occurring on a single version of
the database . therefore , some issues could arise as each development team
migrates to the standard . this needs to happen relatively quickly .
specific issues :
gfd _ pipe _ meters _ snp - probably need to add fac _ type ( wh , ite , etc ) for
possible use as validation in kx functionality
gas _ deal _ location _ dtl - this table refers to facility number rather than
pipe and meter . facility number is an enronism that no other company
will recognize .
pipe and meter general - should try to avoid using pipe _ cd as key . this
value needs to be updateable as pipelines are bought and sold .
- data for facilities is heavily dependent upon enron global facilities
database in current design and functionality - this needs revisiting as
the decision surrounding global is made
cp _ addr _ vw - this table references only internal _ cp _ id and contact _ type _ cd
in conjunction with address _ id . addresses are specific to
internal _ cp _ id , product _ cd , deal _ nature _ cd , contact _ type _ cd , and region _ cd .
need to add deal nature , product , and region to ensure correct address
usage since many companies align their business along these determinants
and addresses may vary
counterparty - there is a mixture of the usage of global counterparty . some
areas indicate a certain amount of independence from enron ' s global
counterparty system , by having a table to capture commodity logic information
on a counterparty such as phone numbers , credit ratings , etc . this would
position cl to become independent at a later date . yet , commodity logic
functions as a subset of enron networks and gcp must make entries to indicate
the usage of a customer by enron networks to obtain an sap id and
support payment processing through to sap . so , cl could not become
independent without further functionality or process changes . so why
not put the added data requirements within gcp to start with ? if
global moves to commodity logic , then this design needs to be revisited for
sure . there should probably be some standardization between
dependence / independence whether or not cl separates from enron .
common data - status is not included in the views being utilized by the
applications . i hope the views have been filtered for active status
only . show was going to check on this .
- concept of mapping others ' codes to ours for processing , is not
supported anywhere in these tables . perhaps that has been handled in
an isolated manner in the fix design ? this will have to be there for internal
release as well as external release . this is critical to the hub
concept .
i look forward to sitting in on your meetings surrounding these issues . let
me know if you have further questions .