Hi, Your understanding is correct. The overall throughput should grow if you scale horizontally.
However in case of distributed transactions many factors can be a reason of why the performance doesn't grow significantly: - the same locks are held by multiple pessimistic transactions; - CacheConfiguration.setMaxConcurrentAsyncOperations is reached; - transactions are mostly executed on a subset of nodes; - etc. In general also you may want to use optimistic transactions that work faster in some cases. https://apacheignite.readme.io/v1.6/docs/transactions#optimistic-transactions -- Denis -- View this message in context: http://apache-ignite-users.70518.x6.nabble.com/Increase-transaction-throughput-with-nodes-tp5411p5485.html Sent from the Apache Ignite Users mailing list archive at Nabble.com.