Re: version/source control in datawarehouse project
- From: "Jeje" <willgart@xxxxxxxxxxx>
- Date: Mon, 20 Aug 2007 20:16:56 -0400
first,
you can manage your SSIS, SSAS ans SSRS files under a source control system like visual source safe very easily. so, there is no issue here.
second,
you can use configuration files to store your database connections. each station need this config file located on the disk and setup for the developper env. (the config file is an xml file which can be edited by each developper to reflect the right servers names)
make sure you use a folder like c:\myproject\ssisconfig which is created exactly the same way on all desktops.
(there is other options to store and share the connections, but this one is easy to implement)
third,
for the version of the databases...
you have to create a good process to manage SQL script files to keep everybody updated correctly. for example, organize this into folders:
/Version1
/Patch1.1
/Patch1.2
....
each folder will contains SQL scripts which modify the database and each developer as to apply the patches on his own environment.
Visual Studio for Database Professional can do this for you.
But I recommend that only 1 person will be in charge of managing the database and the changes.
http://msdn2.microsoft.com/en-us/teamsystem/aa718807.aspx
and finally... communicate! :-)
"YesGoGoGo" <YesGoGoGo@xxxxxxxxx> wrote in message news:1187628671.026950.145670@xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx
Hi!.
I'm a freshman in a datawarehouse company.
All my coworkers are on-site in user side,such as banks,
and our development machine is also given by user.
There is no server or desktop in our company.
After I join this company,I find there's no version/source control in
our company's projects.
Because I used to work in a web-based system development company,
I think it's VERY important to have code under version control.
In my new company,there are 1~3 members in each project.
Everyone of them will terminal to server,open the same project and
modify it.
I try to think if there is a better way to improve this develop cycle.
First of all,I have to have a version control server,such as
subversion.
Because we don't have a server in our company
,maybe installing SVN on development server is a good way.
When sb wants to modify the code(SSIS/SSRS),
he has to update his working copy to the latest version and modify it.
But for me,there is a problem
take SSIS for instance, how do we design datasource connectiion to
make it available for every developer?
if datasource is DB connection,I can change "localhost" to servername
or ip to ensure connections on everyone's computer works.
but for development server,would using servername instead of
"localhost" make db connection lower?because it has to parse the
servername? this is my first convern.
If datasource is file connection,there is a similar situation,using
relative path could make the code fit everyone's working environment
but maybe inefficient when reading file?
In the begining,I think maybe we can copy DB or files to developer's
machine, but I find DB and files are too large(too much data...)
And there are many abosolute path in our code
So if I want to bring version/source control into this project,it
seems we have to modify those path.
After ETL,we have SSAS,SSRS to do.
I want to know if there is a better way or best practice to design the
version control of the datawarehouse project?
For me,it seems like a mess now...
Need some advice
Thanks a lot!!
- Follow-Ups:
- Re: version/source control in datawarehouse project
- From: YesGoGoGo
- Re: version/source control in datawarehouse project
- References:
- version/source control in datawarehouse project
- From: YesGoGoGo
- version/source control in datawarehouse project
- Prev by Date: version/source control in datawarehouse project
- Next by Date: Re: Design choice in the Use of SQl Server Views for ETL
- Previous by thread: version/source control in datawarehouse project
- Next by thread: Re: version/source control in datawarehouse project
- Index(es):
Relevant Pages
|