how stable are the integrations with Cassandra and Hazelcast? Are they ready for productive use?
From the Cassandra integration I read, that it was only a prototype (https://tech.zalando.de/blog/camunda-meets-cassandra-at-zalando/) Is this the current state, or is it outdated?
What is to do to get one of them (Hazelcast or Cassnadra) ready to use? Is there a list of supported or missing features?
There are prototypes for integration of Camunda with Cassandra and Hazelcast. The most mature is the community extension for Cassandra. However, most mature is a relative judgement. It works for some scenarios and for others it does not. The Cassandra extension github project gives some details on how it works and what can be used.
Camunda 7 will unlikely ever work completely on databases like Cassandra since it requires ACID transactions and querying capabilities that relational databases provide best.
I found Jakob Freund’s slides shown at BPMCon2016, stating:
Camunda morgen
…
Unendliche horizontale Skalierbarkeit
Keine Abhängigkeit von relationalen DB
…
Could you provide some more details about you mid-term plans / roadmap concerning horizontal scalibility and “official” support of NoSQL databases (like cassandra) and maybe migration-strategies?
Unfortunately I cannot, other than that we are working in this area. As a company, we haven’t made any more public statements in this respect yet. We’ll announce it in our team blog once that happens. If you are interested in this topic, you can contact our sales department that might be able to share more details in a private environment. For that, see https://camunda.com/about/contact/.