Database Consulting
Customer success story
WorkJam slashed Google Cloud costs through Cassandra database optimization.
Modernizing WorkJam's Cassandra environment helped drive massive operational savings.
Pythian right-sized the over-provisioned Cassandra virtual machines and deleted redundant storage snapshots that were consuming resources.
WorkJam’s Cassandra environment was significantly over-provisioned, leading to excessive Google Cloud Platform (GCP) costs and a lack of internal visibility into resource utilization. They turned to Pythian for specialized expertise to audit their environment and identify cost-saving opportunities. Pythian delivered a comprehensive health check and architectural modernization, rightsizing virtual machines and cleaning up redundant snapshots. This strategic intervention not only reduced monthly spend but also improved long-term sustainability and performance. Additionally, reducing the data size on clusters helped lower latency and improve overall environmental scalability.
Pythian helped us realize we had too much compute power for what we needed. Their expertise in Cassandra allowed us to decrease our instance sizes and save a significant amount of money.”
Director of Operations
WorkJam
$120K
Saved from snapshot cleanup
100%
Visibility into resource utilization
0
Downtime
Eliminate hidden infrastructure waste and drive savings with Cassandra expertise.
The internal team struggled to determine if their powerful machine configurations were optimal for their specific Cassandra workloads.
Pythian conducted a deep-dive technical assessment of the Cassandra environment to identify performance bottlenecks and cost leaks.
Uncertainty regarding database resource requirements
WorkJam lacked the specific technical knowledge to determine if their Cassandra nodes required the high-cost compute power they were currently paying for.
Excessive costs due to unmanaged snapshots
A lack of lifecycle management led to the accumulation of thousands of unnecessary snapshots, costing the business over $100k annually in "ghost" storage.
Mismatched storage tiers for data needs
Standardized machine templates utilized premium disk tiers for small volumes of data, resulting in a significant price-to-performance imbalance.
Lack of specialized database expertise
Their previous partner ceased managed service support, leaving a critical void in specialized guidance for their complex data stack.
The root causes of Cassandra performance stalls and over-provisioning were uncovered.
Pythian performed a technical assessment to establish a baseline for the Cassandra environment's health and efficiency. This involved analyzing node performance against current GCP spend to find discrepancies. Our experts identified that the "huge" environment was significantly larger than required for the actual workload.
Database modernization optimized Cassandra performance at a fraction of the cost.
Pythian updated the environment by changing the machine types to the latest GCP versions, which offer better efficiency. By decreasing the machine sizes while moving to newer generations, we maintained performance and drastically lowered the hourly cost of the instances. This ensured the compute power finally matched the business's actual needs.
Establishing strict data retention policies rooted out hidden infrastructure waste.
Pythian reviewed the entire snapshot schedule to identify where automation had created unnecessary bloat. We worked with the client to define a strict retention period, successfully deleting thousands of old snapshots that no longer served a business purpose. This addressed a major hidden cost in their monthly GCP billing.
.svg){kind=small}
.svg){kind=small}