20.8K ·GitHub
What is this?
An Apache Top-Level Project positioned as 'Database Plus': a layer built over your existing heterogeneous databases that adds distributed SQL - sharding, read/write splitting, encryption - without making you replace them.
An Apache Top-Level Project positioned as 'Database Plus': a layer built over your existing heterogeneous databases that adds distributed SQL - sharding, read/write splitting, encryption - without making you replace them.Is it a library I plug into my app, or a separate server I deploy?Both, by design: ShardingSphere-JDBC is an enhanced JDBC driver (a JAR inside your Java app), ShardingSphere-Proxy is an independently deployed transparent database proxy, and the two can run in a hybrid deployment.
Both, by design: ShardingSphere-JDBC is an enhanced JDBC driver (a JAR inside your Java app), ShardingSphere-Proxy is an independently deployed transparent database proxy, and the two can run in a hybrid deployment.How do I get started with it?For Java, add ShardingSphere-JDBC as a JAR dependency - zero extra deployment. Otherwise deploy ShardingSphere-Proxy as a standalone server and point any MySQL or PostgreSQL client at it. Full docs are on the official website.
For Java, add ShardingSphere-JDBC as a JAR dependency - zero extra deployment. Otherwise deploy ShardingSphere-Proxy as a standalone server and point any MySQL or PostgreSQL client at it. Full docs are on the official website.Do I have to be writing Java to use it?Java only if you pick the JDBC end. ShardingSphere-Proxy speaks the native MySQL and PostgreSQL binary protocols, so clients in any language connect to it as if it were a regular database server.
Java only if you pick the JDBC end. ShardingSphere-Proxy speaks the native MySQL and PostgreSQL binary protocols, so clients in any language connect to it as if it were a regular database server.Any gotchas I should know about before using the JDBC driver?The JDBC driver's ANTLR parser caches parsed SQL, and with many SQL templates the cache grows and can exhaust heap memory (OOM). The README advises setting a reasonable -Xmx heap size for your application.
The JDBC driver's ANTLR parser caches parsed SQL, and with many SQL templates the cache grows and can exhaust heap memory (OOM). The README advises setting a reasonable -Xmx heap size for your application.Can I use ShardingSphere-JDBC and ShardingSphere-Proxy together?Yes - hybrid architecture is a first-class mode. Both ends are wired through the same registry center with a unified sharding strategy, so high-performance Java OLTP apps can go through JDBC while Proxy serves heterogeneous languages, OLAP workloads and DBA management.
Yes - hybrid architecture is a first-class mode. Both ends are wired through the same registry center with a unified sharding strategy, so high-performance Java OLTP apps can go through JDBC while Proxy serves heterogeneous languages, OLAP workloads and DBA management.How mature is the project?It graduated to an Apache Top-Level Project on April 16, 2020, and the README says it has been adopted by 19,000+ projects worldwide. Version 5.5.4-SNAPSHOT is under active development, with 5.5.4 bringing multiple security enhancements and performance optimizations.
It graduated to an Apache Top-Level Project on April 16, 2020, and the README says it has been adopted by 19,000+ projects worldwide. Version 5.5.4-SNAPSHOT is under active development, with 5.5.4 bringing multiple security enhancements and performance optimizations.What does the pluggable architecture mean in practice?A micro-kernel with three decoupled layers: a core layer (query optimizer, distributed transactions, execution engine), a feature layer (sharding, encryption, masking, audit, shadow database) and an ecosystem layer (database protocol adaptation, registry integration). Features combine on demand - the README compares it to building with LEGO blocks.
A micro-kernel with three decoupled layers: a core layer (query optimizer, distributed transactions, execution engine), a feature layer (sharding, encryption, masking, audit, shadow database) and an ecosystem layer (database protocol adaptation, registry integration). Features combine on demand - the README compares it to building with LEGO blocks.What are similar projects?vitess, yugabyte-db, incubator-seata, citus, cockroach, starrocks, juicefs, awesome-bigdata
vitess, yugabyte-db, incubator-seata, citus, cockroach, starrocks, juicefs, awesome-bigdataWhich topics is it in?bigdata, data-pipeline, distributed-database, scalability, cloud-native, mysql, sql, observability, distributed-systems, postgresql, proxy, java
bigdata, data-pipeline, distributed-database, scalability, cloud-native, mysql, sql, observability, distributed-systems, postgresql, proxy, java