I am tasked with creating 3 event chains (A,B,C) for 3 different parts of an ETL load. The 3 parts are all for the same database so the chains cannot run at the same time. Chain A and B need to run Tuesday through Saturday. Chain C needs to run Monday through Friday.
Since I cannot have them run at the same time I need them to kick off each other on days where multiple run (all chains will start at 7am on their various days).
So far I did this Chain A - Schedule Tuesday - Friday. Has chain A, B, and C event chains in it. Chain B - Saturday - Has chains A and B in it. Chain C - Monday - Has chain C in it.
Here is why this didn't work. I took all 3 chains and put them into one chain for the days they run together. I didn't realize (still learning would love some links to learning resources) that if I put a node inside a chain that is the start of another chain, it kicks off that chain. So this morning I had all 3 chains kick off when only chain A should of.
I can think of 3 things that might work, but both have their issues.
I change chain A so it just runs the first node of chains B and C. The issue with this is since I need chain B to finish before chain C, how do I do that if I'm only calling the 1 starting node in chain A? I've only been using the "success" option, is the "completed" option meant for this?
I keep the chains the same but change the nodes in chain A that start chains B and C so they don't kick off those chains. I know SQL Sentry does an _1 or _2 when you do multiple nodes in a single chain. Is this doable?
Somehow create one giant chain with multiple starting points. Least sure of this one.
Any help is appreciated!