What do I need to do in order to manage tracking tokens correctly if I need to use another NoSql database for view projections? I am finding that it will be very difficult to get approval to use mongo in our environment. I could possibly use postgres and maybe with it’s support of JSON I can also skip the complexity of the ORM.
What are people’s thoughts on this topic?
Cheers!
Troy
Hi Troy,
Using PostgreSQL’s way of dealing with the JSONB type is quite helpful I must say; have been using it at clients for some time now.
What I have also done in the past is having the Query Model be a serialized JSON, stored in a String with some added identifiers for simple queries, and keep that in any RDBMS really (I believe I used MySQL at the time).
A part from that though, I have also with relative ease made PoC’s in the past where I used ElasticSearch or Neo4J as the Query Model, whilst the TokenStore was still maintained through the JpaTokenStore.
Any how, it’s doable, but I understand you’d like to keep the tokens closer to the NoSql solution you’re going for.
That’s my 2 cents to the situation.
Cheers,
Steven
Thanks for your thoughts on this matter. I think I lean toward postgres and jsonb.