Handle more than 150,000 transactions per second. Using the Actor model for concurrent transaction processing and RocksDB for durable storage.

The approach to updating outdated data while preserving determinism in a total ordering architecture

Simple Java implementation of the Raft consensus algorithm for fault-tolerant, distributed command replication. Supports UDP-based node communication

Personal Mathematical Notes

Embedded database appending 180,000 records/second with low latency (using for Raft)

A simple high-performance ring buffer handling 10,000,000 messages/sec.

Handle approximately write 60,000 RPS with latencies consistently under 15 microseconds. Utilizing gRPC, ring-buffer and robust leader.

Maven central namespace

Since Agrona collections do not support a ring buffer with multiple consumers, this repository provides a simple implementation of OneToManyRingBuffer and ManyToManyRingBuffer.

Handle more than 200,000 write-requests per second with Lmax disruptor, Kafka and gRPC

Append nearly 30 million commands per second.