Pepperdata - Complete Spark Cluster performance monitoring
What do you like best?
Pepperdata provides insightful details and helpful metrics to monitor cluster performance. The user-friendly interface helps to navigate several details like CPU, Memory, Storage usage, I/O etc.
Pepperdata also helps to optimize the cluster settings to ensure enhanced resource management.
App spotlight feature helps to figure out slow jobs or any data skew scenarios
What do you dislike?
Pepperdata can integrate automated alerting in case of any cluster parameter shows alarming behavior. It can also provide features like AI-based detection of resource containment and alert accordingly.
It will be good if Pepperdata can integrate workflow managers like Oozie in the monitored services and map Oozie coordinator/Job Ids to spark Application ids. This will help in easy mapping between interfaces and will result in quicker root cause analysis of issues.
Recommendations to others considering the product:
Pepperdata is a helpful spark cluster monitoring tool, which provides all relevant metrics of a cluster in a centralized portal.
What problems are you solving with the product? What benefits have you realized?
1) Automated Spark cluster optimization
2) Root cause analysis of any issues with Spark jobs
3) Detecting data skew scenarios
