cancel
Showing results for 
Search instead for 
Did you mean: 

Automatic end to end deployment

Former Member
0 Kudos

Hi All,

We are planning to build an automatic end to end deployment using API and CLIX. Though CTS+ has given some flexibility in systematic deployment, we have a requirement to completely automate MDM deployment process.

For eg: If I'm splitting a coupled numeric field into 2 seperate fields as part of the release then I need to:

1. Create Syndication Map in Dev to take back up

2. Transport the Syndication Map only to QA: Can be done either via Export / Import or Repository Schema

3. If done via Schema, the Repository Schema import needs to be automated to Accept the changes.

3. Execute the Syndication after map is moved

4. Do schema changes to Dev repository

5. Create Import Map to re-import the data in new fields

6. Transport Repository Schema to QA

7. Execute Import Map

The question is how do you build this? Esp. export & import repository using API's. Would appreciate if you can list down any specific commands as well that can be used.

Thanks!

Accepted Solutions (1)

Accepted Solutions (1)

former_member205403
Active Contributor
0 Kudos

Hi,

looks a very interesting requirement. I never tried it myself, but you can Try by following below mentioned steps:

1. Share archive directory between DEV and QA.

2. Create Syndication Map in Dev to take back up of data.

3. Define Import Map in Dev repository to import data in new fields.

4. Transport the Syndication Map only to QA and take dat.a backup. Also I would suggest to prepare an archive file of QA repository, just in case things goes . This can be done using below mentioned CLIX command.

CLIX cpyArchive MDMHostSpec RepositorySpec [-A filename]

In your script you case define filename as with MDM system ID and date time stamp. (just to differentiate between dev and QA rep archives)

5. Once Back up is done execute below mentioned command from your deployment script.

CLIX cpyArchive MDMHostSpec RepositorySpec [-A filename] [-XS]

-F option will be used to overwrite existing repository...

where -XS option is used to create a schema only archive (.a2a file).

6. Again using Clix command unarchive this .a2a on QA server and overwrite it on existing QA repository using command

CLIX cpyUnarchive MDMHostSpec RepositorySpec filename [-F]

**File name you can find based on time stamp.

7. Once repository is unarchived then in QA, then import Data back from the file created in step 2.

As said earlier, I have never tried it myself and you may face challenges. Also I am not sure if import map defined in step 3 in dev will be available in QA or not. But you can check on this and revert with the result.

For more details about clix commands please refer Console reference guide.

-


But one problem which i can think of in this solution is, your users in QA may disappear... but you can manage this by downloading User's list from database in a file with password, and then after deployment uploading them back using database command..

Also I would like to add, above solution could be one possible solution but no guarantee that it is the best solution.

Best Regards,

Shiv

Former Member
0 Kudos

Thanks Shiv.

I dont think moving an archive from DEV to QA will be a good ideas as later the same process needs to be followed from QA to PROD. Hence, I wnated to know how can we transport Repository Schemas from one environment to another using wither API or CLIX. If we can get that automated, then I suppose we can work things out for executing maps etc.

Also, whats the difference between a "Schema only archive" vs "Repository Schema". One would create .a2a file & another xml, but "Repository Schema" gives options to Accept / Reject.

Thanks,

Ketan

Edited by: Ketan Phanse on Feb 25, 2010 3:18 PM

former_member205403
Active Contributor
0 Kudos

Hi,

you are right... But I don't think you can automate schema export and import process using Clix.

Regards,

Shiv

Answers (0)