Scylla Manager

NoSQL database management for Scylla scale-out clusters

NoSQL Cluster Management System

Automate Repetitive Tasks

Operations teams can quickly schedule recurring tasks such as node rebuilds and repairs across their cluster. A single instance of Scylla Manager is able to manage NoSQL clusters spread across multiple regions and datacenters.

Disaster Recovery

Scylla is designed for high availability in case of node failures but systemic disasters can always strike. Use Scylla Manager to quickly restore your data from an incremental or complete backup after an outage, minimizing unnecessary downtime. Rapid disaster recovery makes Scylla one of the most resilient NoSQL systems for big data management.

NoSQL Cluster Monitoring

Scylla Manager aggregates performance data across the cluster and makes it available in Prometheus format. Automatic topology discovery using node agents simplifies setup and NoSQL database management. Deploying Scylla Monitoring Stack alongside Scylla Manager empowers operations teams to track real-time metrics using out of the box Grafana dashboards.

Faster Troubleshooting

Historical and real-time charts help teams troubleshoot issues related to metrics such as latency, request timeouts, and cache hits. Quickly identify real-time database loads and currently running background processes. Teams are able to easily find and fix issues that may require table repairs, node rebuilds, or instance restarts to maintain service level agreements (SLAs).

Technology

Scylla Manager is a lightweight stateless Go application that uses Scylla as its persistent store. A Scylla Agent runs alongside each Scylla database on the node to collect local database metrics/information and send them to Scylla Manager over HTTPS. Operations teams can centrally interact with all nodes of the Scylla cluster through Scylla Manager’s sctool.

Licensing

Scylla Manager is free to use for all customers up to 5 nodes for clusters built with Scylla Open Source. Customers using Scylla Enterprise can connect an unlimited number of nodes with clusters deployed across multiple datacenters.