All Products
Search
Document Center

FAQs

Last Updated: May 29, 2019

Q: What is Alibaba Graph Database (GDB)?

Alibaba Graph Database (GDB) is a real-time, reliable online database service that supports property graph model for efficiently storing and querying highly connected data set.

Q: When should I use GDB?

  • Traditional Rational Database service can’t provide effective query optimization especially when dealing with highly connective datasets. GDB specifically designed this scenario to provide an order of magnitude higher of capabilities in query the multi-hop or recursion data thus is suitable for digging its implicit connectivity information
  • Take the Knowledge-Graph for example. GDB provides the native graph abstraction which is perfectly match the property graph model, thus user won’t care about how data is organized in rows and tables of traditional databases.
  • Last but not least, typical user needs the real time processing besides the capabilities above. GDB also provides the special optimization on online operations with ACID transaction support. With this user can also get benefited in both TP/AP scenarios.

Q: How does GDB work?

GDB uses TinkerPop Gremlin query language for its standard graph service, with its inner design based on the property-graph model abstraction. It employs the common master-backup mode to ensure its availability, and triple copies of datasets as an enhancement. Mutiple backup nodes (read only) can be introduced as parallel computing nodes. GDB also provides feasible optimization for online services with full ACID transaction support. Also, user is able to imgrate its services to GDB easily via OSS import service, or with direct write.

Q: Why is GDB better than many traditional graph systems?

Compared to traditional graph systems, such as Neo4j, JanusGraph or other open source graph systems, GDB is considerably more efficient and easy to use. GDB’s benefits:

  • Standard graph query language: Supports property graph model and is highly compatible with TinkerPop Gremlin query language.
  • Highly optimized engine: Highly optimized graph computing layer and storage layer, which supports fully ACID transactions.
  • High availability: Automatically detects node failure and replaces failed nodes. Offers master-slave dual node backup architecture without service interruptions.
  • Multiple maintenance features for cost efficiency: Provides rich maintenance features such as backup/recovery, automatic upgrade, monitoring, failover etc.