High-Concurrency Shared-Disk Active-Active RDBMS
Architected from the ground up in Rust for symmetrical multi-master clusters. Scale transactional and analytical workloads across unified shared storage with drop-in enterprise driver compatibility and zero application rewrites.
# Enterprise client connects to active-active cluster service:
$ ringlock-sql admin@ringlock.io:1521/prod_service
Ringlock Interactive Shell: Release 1.0.0 - Production on Mon Oct 26 2026
Connected to:
Ringlock Active-Active Cluster Engine 1.0 (Rust / x86_64-unknown-linux-gnu)
Cluster Topology: 3 Active Nodes | Shared Storage: NVMe-oF / RDMA | Cache Fusion: GES v2
SQL> SELECT instance_name, host_name, status, active_state FROM GV$INSTANCE;
INSTANCE_NAME HOST_NAME STATUS ACTIVE_STATE
------------- --------------- ------ ------------
ringlock_1 node-01.ord OPEN NORMAL
ringlock_2 node-02.ord OPEN NORMAL
ringlock_3 node-03.ord OPEN NORMAL
3 rows selected. Elapsed: 0.000351 seconds.
Shared-Disk Active-Active
True symmetrical multi-master clustering. Multiple nodes mount the same shared block volume concurrently over NVMe-oF and kernel-bypass RDMA without primary-replica failover latency.
Cache Fusion & DLM
Global Cache Service (GCS) and Global Enqueue Service (GES) transfer dirty 8KB memory blocks across nodes at interconnect wire speeds, synthesizing Consistent Read (CR) clones on demand.
In-Place Slotted Storage
Zero-vacuum, in-place update engine with rollback Undo segments. Prevents the table bloat, index fragmentation, and write amplification common to append-only architectures.
Enterprise Client Compatibility
High-performance C-ABI shared library and Java JDBC driver shims allow enterprise data layers, legacy connection pooling, and client binaries to connect without code changes.
Vectorized HTAP Query Engine
Combines sub-millisecond transactional OLTP with vectorized Apache Arrow and Apache DataFusion execution for ultra-fast analytical queries on unified shared storage.
Memory-Safe Rust Engine
Engineered without garbage collection pauses, memory safety vulnerabilities, or legacy technical debt. Validated under multi-node chaos simulation and continuous automated testing.