Querying and writing data from a distributed database such as Apache Cassandra can be tricky at times. Often, write-heavy usage patterns can cause performance bottlenecks, and force applications to wait for operational completion. With the DataStax Java driver, both read and write operations can be executed in an asynchronous, non-blocking fashion.
A common problem with periods or bursts of high-write throughput is that the nodes can become overloaded with traffic. When this happens, it is usually because data is being written faster than Cassandra can keep up with. It typically results in a plethora of problems including high CPU, dropped requests, connection timeouts, constant compaction, and system resources so ...
