I have had to go to manually unscheduling the ETL when a weekend arrives where there is scheduled maintenance on a target DB. If I don't do this, the ETL will fail and abort -- forcing me to reset or recompile in order to run it again the next time.
I am looking for a way to build a job that could be included in the sequence so that, if the connection attempt failed on a source DB, I could trigger a path that would be considered a non-failure from the standpoint of the sequence. The idea is that I could leave the ETL scheduled all the time and, if it attempted the connection and failed, it would simply log the failure and just take a path that terminated the sequence. The sequence would consider itself to have completed successfully so the next scheduled time to run, it would just do so without having to be reset or recompiled.
Anyone have ideas? I'll keep working on this but just wondered if someone had already conquered this hill.
![Smile :-)](./images/smilies/icon_smile.gif)