Questions? Feedback? powered by Olark live chat software

Continuous Query Cache

Feature of Enterprise HD, and Enterprise

Queries with Zero Latency

Continuous Query Cache maintains the result of your filtered queries that you perform on a Hazelcast Map. You can think of it as a queried Hazelcast Map which is local to your application. Along with the query result, this cache gets a continuous stream of events generated by the changes within your map. By this way, the cache offers always up-to-date query results in a real-time manner, meaning that your queries on the map will have a zero latency.

You don’t need to run any distributed queries on your cluster to get the desired data; relevant data will come to your application local automatically. All operations on the Continuous Query Cache are performed locally like a java.util.concurrent.ConcurrentHashMap, so it is fast.

Since it maintains an up-to-date local copy of a specified subset of your distributed map, Continuous Query Cache is beneficial in a lot of applications whether they are server or client based.


  • For the applications that involve repeated queries and that require instant access to the query results.
  • As a way to access and use the query results using NamedCache API.
  • As an entity similar to a Near Cache since it has an up-to-date set of data local to your Hazelcast member or client. The difference is that a Near Cache is based on invalidation, but Continuous Query Cache maintains always current data.
  • As a building block for complex event processing systems.

Key Features

  • Access either from a Hazelcast member or a client.
  • Run queries with indexes.
  • Perform event batching and event coalescing.
  • Specify a maximum size so that the entries can be evicted.
  • Listen to events using Hazelcast’s map listener.
  • Get streaming events in the same order as they are generated on the map entries.