I've been looking into building a data warehouse, and I'm not sure it will make my life easier!
I would like to just come up with my "dream schema" and run update scripts on a daily or event-driven basis, but there seem to be a million concepts related to data warehousing: data marts, dimensions, star schemas, snowflakes, EAV tables, and so on. It seems more straight forward if I could just connect directly to the operational systems. I wonder if all that theory is really necessary for a basic data warehouse. Maybe I should just call it a "centralized/ redundant DB." On Sep 21, 6:18 pm, mdipierro <mdipie...@cs.depaul.edu> wrote: > I do not know. I can ask but I suspect the former. > > On Sep 21, 6:54 pm, snfctech <tschm...@sacfoodcoop.com> wrote: > > > Thanks for your reply, Massimo! > > > Out of curiosity, are they using migration scripts to create/ update > > thewarehouseDB, or a middle-ware product like Talend or Pervasive? > > > On Sep 21, 3:35 pm, mdipierro <mdipie...@cs.depaul.edu> wrote: > > > > At the university where I work they run many different systems > > > including peoplesoft with oracle. Neverthess they find it convenient > > > to have a datawarehouseand I think it is a good idea. This allows > > > other users to get access the data on a need to know bases and allows > > > them to manage permissions. Moreover if any of those systems gets > > > upgraded you only may need to change the query to move data to the > > >warehouseand not the third party programs that depend on those data. > > > > Even if you can connect to multiple databases your life will be easier > > > with a datawarehouse. > > > > On Sep 21, 4:38 pm, snfctech <tschm...@sacfoodcoop.com> wrote: > > > > > The big reason I started investigating web2py was multi-db > > > > connectivity. > > > > > I've got a large dashboard application I'm about to start work on that > > > > has data on MSAccess, MySQL, Pervasive, FileMaker and flat file DB > > > > engines. This all needs to come under one roof and interaction needs > > > > to be seemless to the user. For example, I'll need to have an order > > > > creation form with relations to separate product, member and employee > > > > databases. > > > > > Before finding web2py, I was told the best way to do this would be to > > > > create a batch updated datawarehouse. I initially thought that was > > > > an unnecessary step - why not just connect directly to all your data > > > > sources, if you have that option? > > > > > But then I found one possible advantage of the datawarehouse- > > > > enabling "middle tier" development staff with tools like OOBase (which > > > > cannot make multiple connections). Table creation and write > > > > permissions would be managed for those users, of course. > > > > > So, what do you think? Any other reasons for creating a batch updated > > > > datawarehousewith a unified schema? Or are direct connections to > > > > multiple DBs the way to go? > > > > > Thanks. --~--~---------~--~----~------------~-------~--~----~ You received this message because you are subscribed to the Google Groups "web2py-users" group. To post to this group, send email to web2py@googlegroups.com To unsubscribe from this group, send email to web2py+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/web2py?hl=en -~----------~----~----~----~------~----~------~--~---
