Chartboost Case Study (PDF)
Chartboost Experiences Significant Amazon Web Services (AWS) Savings
“The throughput improvement is worth the cost of Pepperdata,” says Clubb, “We installed on the production cluster and immediately saw performance gains—what’s better than that?
We immediately saw a performance boost and we could measure that easily. We’re very happy “
– David Clubb, Chartboost Senior Data Engineer
Chartboost faced managing explosive data growth. They had no way to effectively monitor and troubleshoot, or view granular health and status metrics for all of the clusters from a single, unified dashboard. Chartboost also had no way to effectively manage a multi-tenant cluster and ensure that high-priority jobs completed on time.
Platform Spotlight enabled Chartboost to gain deep visibility into the use of CPU, memory, disk I/O, and network resources by container/ task, job, and user on a second-by-second basis to troubleshoot the problem. They quickly determined that a single, large job was consuming significant resources and causing nearly 100 percent CPU utilization multiple times a day. Capacity Optimizer immediately improved cluster throughput by 31%. Pepperdata enabled Chartboost to configure policies to prioritize specific users, jobs, or queues, and ensure that high-priority jobs completed on time, regardless of workload contention.
Pepperdata Platform Spotlight enabled Chartboost with the visibility to pinpoint and troubleshoot the problems in their cluster, enabling Chartboost to reduce its total AWS node count by 30 percent.