-
Hello, our team has been evaluating this product within a kubernetes cluster where we want to scale a helidon microservice to multiiple replica pods using a microstream file store in shared PV. I was a bit surprised when updates to one pod did not propagate to the other pods and eventually found this one section in the documentation that talks about the limitations of a shared Microstream database: https://docs.microstream.one/manual/storage/faq/file-storage.html My interpretation of the last paragraph is that if we wanted to use multiple pods, we also would have to create copies of the Microstream database (each pod with its own) OR we would have to create a data master microservice that acts as a gateway that the worker processes can call. I feel that each solution kind of defeats the main advantages of microstream, which is simplicity. Can you share in more detail (via examples) what the best practice is on how to architect a microstream solution in a K8s cluster? Thanks, |
Beta Was this translation helpful? Give feedback.
Replies: 1 comment 2 replies
-
Hello Derek, Your interpretation of the documentation is correct. Best regards |
Beta Was this translation helpful? Give feedback.
Hello Derek,
Your interpretation of the documentation is correct.
Currently Microstream does not provide the necessary features for replication and synchronization of several nodes.
We are currently working on all those features to provide Microstream – Cluster solution. But we can’t tell when this will be available and in which way it will be released.
Regarding the DTO mapping: It is possible to use Microstream as a serializer. A simplified API for that will be available with the next release.
Best regards
Your Microstream Team