A drop-in middleware for serializable DB clustering across geo-distributed sites

Abstract

Many geo-distributed services at web-scale companies still rely on databases (DBs) primarily optimized for single-site performance. At AT&T this is exemplified by services in the network control plane that rely on third-party software that uses DBs like MariaDB and PostgreSQL, which do not provide strict serializability across sites without a significant performance impact. Moreover, it is often impractical for these services to re-purpose their code to use newer DBs optimized for geo-distribution. In this paper, a novel drop-in solution for DB clustering across sites called Metric is presented that can be used by services without changing a single line of code. Metric leverages the single-site performance of an existing service's DB and combines it with a cross-site clustering solution based on an entry-consistent redo log that is specifically tailored for geo-distribution. Detailed correctness arguments are presented and extensive evaluations with various benchmarks show that Metric outperforms other solutions for the access patterns in our production use-cases where service replicas access different tables on different sites. In particular, Metric achieves up to 56% less latency and 5.2x higher throughput than MariaDB and PostgreSQL clustering, and up to 90% less latency and 26x higher throughput than CockroachDB and TiDB, systems that are designed to support geo-distribution.

Leader Confirmation Replication for Millisecond Consensus in Private Chains

“…These types of database systems need to maintain data consistency to ensure that all operations are executed correctly and in the same order to all consensus nodes. For instance, CockroachDB [1] and TiDB [2] use Raft [3] to support geo-distributed consistency [4]. Open source private chain systems such as Hyperledger Fabric [5] use leader-based consensus models such as PBFT [6] and Raft for transaction consensus [7].…”