Instantly analyze your Trino (formerly known as PrestoSQL) and PrestoDB clusters to get deep workload-level insights on how your clusters perform, identify heavy users and bottlenecks, and effectively optimize performance and concurrency.
slack icon Join the Analyzer community on Slack Join

Reveal the Secrets Behind Your Presto and Trino Clusters

Extract & Analyze Query Performance

Instantly and seamlessly summarize key query metrics, delivered as summary.jsonl file which can also be easily leveraged by third party applications.

Monitor Cluster Health

Continuous monitor of clusters health and performance by tracking key metrics such as queries elapsed time, data scanned, memory consumed and joins execution.

Identify Query Patterns & Outliers

Reveal query detail patterns, such as query peak memory and input data read by query, identify outliers and extract their IDs for further analysis.

Identify Resource - Heavy Operators

Identify Presto operators and determine whether most resources are spent on scanning tables, joining tables or on other operations.

Optimize Utilization per Table

Investigate table and user resource consumption by time utilization and input bytes read.

Optimize Joins

Identify joins that are performed in the wrong order and which joins operations can be accelerated by converting into broadcast joins.

Collect. Learn. Optimize.

Collect Presto Metrics

The Analyzer collects Presto and Trino workload statistics, and analyzes them. The analysis report provides improved visibility into your analytical workloads, and enables query optimization – to enhance cluster performance.

The Analyzer collects, and stores, QueryInfo JSONs for queries executed while it is running, and any historical queries held in the Presto Coordinator memory. The collection process has negligible compute-costs, and does not impact cluster query execution in any way. Ensure that sufficient disk space is available in your working directory. Typically, a compressed JSON file size will be 50kb – 200kb.
Schedule A Demo

01

The Presto & Trino Analyzer Report

Insights are continuously revised, based on real-time usage and performance:

  • Query detail – query peak memory, input data read by query
  • Table activity – query peak memory, input data read by query
  • Presto Operators – wall time usage, and input bytes read by operator

Sounds interesting? Check out the sample report.

 

Key Insights on Your Presto & Trino Queries

Identify Top Spenders
Understand users' resource consumption patterns, and easily identify heavy spenders to make sure budget is used effectively and according to plans.
Monitor Cluster Activity by Time
This charts demonstrates cluster scheduled time by the hour, which is useful for low-level trend analysis and planning cluster scaling policies.
Identify Peak Memory Used by Queries
This chart helps identify memory-heavy queries that use more than 10GB of memory, which are located inside the purple rectangle. Further investigation is recommended in this case.
Identify Elapsed Time by Queries
This chart helps identify slow queries, which require further optimizations.
Monitor Joins Distribution
Joins are a critical and resource-heavy operator. In this scatter plot each join operator is represented by a dot. The x coordinate is the data read from the right-side table, and the y coordinate is the data read from the left-side table. All join operations that appear below the diagonal were executed in the wrong order a nd should be revised so that the left side table size will be bigger than the right side table.
02
We use cookies to improve your experience. To learn more, please see our Privacy Policy
Accept