Skip to main content

Hi,

It seems that our database and model are out of sync. I have added a new column to a table and used this column in new functionality. When deploying this new version to our test environment, the new column was not part of the upgrade (even while the upgrade box was checked) and subsequentially the execution of related procedures failed.

Although we have to look at the cause I would also like to know if there is a possibility to re-sync (if you can call it as such) the model with the database schema.

Any thoughts?

Depending on the version of the software factory you are using the most likely cause is: 

Older versions where you have projects and manually need to create project versions:  either the change was made without making a new project version first, or the origin version is incorrect. 

Newer versions that work with models/model versions: probably the origin version is incorrect. 
Did you check if the origin version matches the database?

About the resync, there are several ways to fix this problem. But I think looking at the origin version, the source branch, would be your first bet. Assuming you are working with one of the latest versions. Did you set the source to a specific version, or based on the database version?


Hi Erwin,

We are using 2024.1.13. In this situation we have a development (latest) version of 1.12 and a test version (separate environment) of 1.11. When upgrading we use the target version (1.11) as the source version when generating the upgrade scripts. This normally works fine. It looks like an older branch (6 weeks old) being merged might have caused the problem. Normally branches are not more than a couple of days old. I was able to fix it to use one version older (1.10) as the source. Strange because the test environment really was at 1.11 (according to the application model version).


I'm not sure what went wrong here. Like you mentioned the branch is probably the key culprit. Since you managed to fix it I assume we can close this topic. If you have problems like this again then it's wise to investigate this further. But for now it seems to be a human error in the branching/merging/deploying process.