Monitoring Cassandra
Complex applications that require storage of a massive amount of data and a flexible data structure require a new form of database. Cassandra database offers scalability and high availability without compromising performance on commodity hardware or infrastructure. This makes it an ideal platform for mission-critical data and for replicating across multiple datacenters. Cassandra is a best-inclass solution, providing low latency for users and peace of mind, with an architectural rigidity designed to survive regional outages.
Business Challenge
Modern NoSQL databases like Cassandra are designed to support massive data processing and provide an equally large storage capability, but hosting Cassandra clusters with a large number of nodes can greatly increase the complexity of your data infrastructure. Understanding the performance of your Cassandra clusters is critical for diagnosing issues and planning capacity.
Foglight for Cassandra provides comprehensive performance monitoring and administration for all nodes in a cluster from a centralized console. You can collect statistical data from all JVMs in a cluster and key performance metrics like memory utilization statistics, task statistics of thread pools, storage statistics, CPU usage, operational performance, latency, and bottlenecks.
Key Features
Foglight for Cassandra provides a consolidated view of all monitored Cassandra Clusters, containing information on cluster structure, nodes, health status, and other key metrics. Derived calculations provide insight into the overall cluster workload taking into account activity on each node in the cluster.
View Cluster Health at a glance, displaying a topology of the cluster structure or a list of nodes with relevant performance and availability metrics. Tables can be viewed by cluster or namespace, aggregating associated metric information across nodes. Node Information is granular and includes Node Health and Alerts, Availability, Workload, Statements, Connections and Operations.
Understand every component of Cassandra performance. Granular performance metrics such as Reads/Writes, Caches, Bloom Filter, Buffer Pool, Memtables, and Commit Log are collected and stored for historical and trend analysis. JVM information is also collected and analyzed to help ensure optimal configuration and performance.
Understand response times with a breakdown of Client Requests to the node by type, showing metric histories for average latency, request counts, and request errors, broken down by error type. Understand the performance and availability between all nodes in your cluster. Foglight for Cassandra monitors all Connections including dropped messages, timeouts, and pending messages.
Alerts can be generated based on preset rules and thresholds as well as dynamically, based on situational anomalies. Alerting has been designed to help ensure the performance and availability of your Cassandra environment at all times and to keep it performing optimally.
Additional details will be provided shortly.
