Kyuubi supercharges Spark SQL with resource isolation, multi-user support, interactive performance, and high availability.
No more:
- Cluster-wide slowdowns from one heavy query ☠️
- Missing security controls 🔐
- Session crashes without recovery 🧨
Why Kyuubi > Spark Thrift Server?
- 🧩 Isolates sessions & allocates resources per user
- 🔐 Integrates with Apache Ranger for row/column-level security
- ⚡️ Uses ZooKeeper for failover + session continuity
- 🚀 Pre-warms Spark sessions for blazing-fast interactive queries
Key Features:
- Multi-tenancy: Each user gets their own Spark instance
- JDBC/ODBC + REST API support
- Python, SQL, Scala compatibility
- HA with ZooKeeper
- Extensions (UDFs, auth, logging)
- Works great in Arenadata Hadoop (ADH) stacks
Try it out and bring efficiency, security, and scalability to your Spark environment!
Top comments (0)