I would ask some clarification about custom contexts and batch procedure.
I succesfully implemented the canonical event model but i would extend my js tracker in order to pass some custom variables to the clojure collector. I would avoid to set up an iglu repository since the data I need are few and simple.
I modified my tracker following this guide:
Running out of ideas I added a “context” column to my Redshift fat table but after my job executions I can’t find anything in my db. (probably bad events generated)
After some days I tried to follow this guide but also there is not clear where my data will land and which format I should expect. If I may move some critics to official guides, I find them very confusing about this section. After a couple of weeks I can’t understand which implementation could fit my needs yet. Furthermore I can’t understand how a correctly defined custom context will be stored in a Redshift cluster.
What do you think about? Requiring simple key-value variables to record, which implementation would you suggest?