Distributed Data on Distributed Infrastructure
summary
ArangoDB runs as a native multi‑model database on Mesos with a single query language, per‑collection replication factor 2 and shard placement; early integration required 5,000 lines of C++ to manage state but the Mesos SDK now generates most framework code. Cross‑datacenter replication is implemented asynchronously using Kafka MirrorMaker plus a new ArangoSync tool, providing offsite backups and failover (not active‑active), with TLS, Kafka buffering/backpressure, and an agency consensus component that tracks cluster state. The demo adds coordinators and DB servers live and shows shard movement and controlled scaling.
AI-written summary · about
chapters (13)
- 0:00Introduction
- 2:22Why Distributed Data
- 7:08Arango DB
- 8:56Multimodel
- 11:43Deployment Roles
- 12:56Failures
- 13:53Frameworks
- 15:15SDK
- 17:01Replication
- 17:14Data Center Replication
- 20:01Kafka Cluster
- 24:44Replication Across Data Centers
- 27:22Demo
related talks
38mDisaster Recovery with a Distributed Database
43mCreating Fault Tolerant Services on Mesos
39mHandling Billions of Edges in a Graph Database
29mMulti Active-Active and disaster recovery with MongoDB database center
29mEnterprise Scale: How a Distributed Architecture Can Deliver Scale and Resilience
20mDomain Driven Design & NoSQL
32mA New Sharding Architecture for MongoDB
41mHTTP API In Action