h2 DB Sychronization

Question

I'm using h2 version: 1.4.187 and have the following problem:

My architecture: 2 servers, on each 1 application and 1 h2 DB.

The DBs are working in a cluster; that means if one of the h2 DBs is down the other DB will take over. Let's call them A and B.

If both are alive and a user writes something in the application, the data will be replicated on A and B. But if A is down and the user writes something during this time, it will only be on B, which makes sense, but if A is up again, it doesn't synchronize automatically and take the data from B. I'm looking for an automatic sychronization. That means if a db is missing data while it's down, it should take itself independently the missing data when it is online again.

Is there is a solution for that?

Answer 1

H2 currently has a limited form of such high availability.

Another existing tool is H2HA, it works similarly.

In both cases, getting A or B back after a failure is not done automatically. You would have to copy (or rsync) the database file. You could automate this, but there would be some small downtime involved (in the order of seconds typically).