I am seeking information concerning integrating Subversion with STAT. The integration connectivity is established and we have successufly migrated an object archived from Subversion to a tmp location on an EBS instance. However, with that in place, we have found no clear direction on best techniques to configure.
The plan is to store the file types in Subversion such as:
.pks - PL/SQL Package specs
.pkb - PL/SQL Package body
.class - Java Class
.xml - Form
.rdf - Report
.msg - Message
.jsp - JSP
.xml - OAF XML
.sql - SQL DDL
.jpx - JPX
We are planning to create post migration scripts for each file type.
Is there any recommendation concerning the structure of the Subversion repository in relation to the EBS environments?
The question I have is, why would you want to? We were considering using TFS for version control and try to integrate with Stat but, since Stat can handle versioning of pretty much all the object types we need, we shelved that idea. Our only gripe is there is nothing in Stat for versioning documentation so we have to use TFS for that.
Good question and thanks for the reply. I have expressed that point with the the EBS development team, but they want to use subversion as their primary source code repository. My assumption is the interation with subversion with some file types may be easier than with STAT.
I'm sure someone more knowlegable will provide an official answer but, as far as I am aware, there is no integration with 3rd party version control tools. Probably the best (only) way you could do it would be to get the required files from Subversion, place them in your Stat working directory and create an archive set from there based on those objects. Some of our Dev's do this with VSS although we have made it clear that it is no longer required and VSS is out of support.
Yes! Stat does integrate with Subversion. This functionality was released in Stat 5.6. Please see below for a screenshot of the integration indicator.
If you would like a whitepaper on setting up Subversion/PVCS/Perforce/CVS please add a comment or contact me directly and I will get you the information.
Stat Solutions Architect
William R. Hart
Quest Software, Inc.
In addition to Bill's comment, we "strongly" and I mean "stongly", ok, just do it ... recommend to follow the same directory structure as EBiz. For example for .sql files, they are probably in XXCUST/sql or XXCUST/admin/sql. For OAF files, they are probably in XXCUST/oracle/apps/xxx/xxx/xxx or in /oracle/apps/XXX/xxx/xxx.
Does that make sense?
-Fern, Stat PM
Stat can handle versioning of documents. It's just another file type. Can be genapps or generic ebiz files, up to you. You just have to setup file locations and set the format to image. Call me if you have more questions!
Thanks William. Yes, the integration to Subversion is in place and working. Currently we are on STAT 5.6.4. I discussed my questions with Quest support yesterday and this was the take away from the discussion, or at least the understanding I have at the moment.
When integrating with a VCM tool, first attempt to utilize the Predefined file types because STAT should know how to process the objects during a migration to an EBS instance. If the object to store in subversion is not in the Predefined list, then customization is necessary.
My current direction for customization is to create a new appl top location (I will call "svn_top") for the customized file objects, and create this location on each EBS server. I will use the new appl top to define the migration location for the custom objects on the EBS servers. The custom object will need a post migration script (using the Templates/Generate>Define Generate Command option) to handle processing which may include copying a file, converting a file, executing a file, etc.
If there is are other options or information, I would be interested on any advice. Thanks again.
Fern, Thanks for the input. The subversion structure has been one of the questions I have been trying to get a better understanding of.
I am interested with integration of Subversion with STAT... ..would you please send me the whitepaper. Thanks!
Tim, The STAT documentation and Quest support can help guide establish the connectivity. Regarding configuring STAT to handle custom objects, my company is slowly working on the different custom object types script processes. We still have much work to do.
Yes - I will private message you the whitepaper on the Subversion Integration.
We are working through our STAT implementation with SVN and we are finding the integration works well as long as we don't bring code branching into the equation. Can you provide any suggestions on how we can configure the tool to allow us to utilize this? In our current SVN repositories we are using the following structures:
The drawback we are finding when trying to integrate into STAT is that we can only link to a single bucket. We would like to be able to utilize different structures above to move code between our DEV, SQA, and PROD environments. Any insights would be appreciated.
Could you please send me the whitepaper as well. I need assistance on the integration between STAT and CVS. Thank you.
Absolutely. I will also see about getting that document posted as public on this community.
In reply to William.Hart: