Using query recommendations in Hybrid Manager

Query recommendations in Hybrid Manager help you continuously improve the performance of your database workloads.

By correlating detailed query diagnostics with actionable recommendations, Hybrid Manager enables you to:

• Identify which queries are slowing down your systems.
• Understand how to optimize those queries.
• Take action to reduce query latency and resource consumption.

Who uses query recommendations?

Query recommendations are valuable for:

• DBAs and platform engineers tuning database performance.
• Data engineers optimizing data pipelines and batch jobs.
• Developers diagnosing application performance bottlenecks.
• SREs and observability teams monitoring database health across environments.

What are query recommendations?

Hybrid Manager collects in-depth query diagnostics and analyzes them. The platform identifies opportunities to optimize your workload, including:

• Index recommendations to improve data retrieval performance.
• Statistics recommendations to improve query planning and execution accuracy.

Why use query recommendations?

Query recommendations help you:

• Improve query performance with minimal manual tuning.
• Focus tuning efforts on queries that will provide the highest impact.
• Continuously adapt to evolving workloads — as query patterns and data volumes change, so do your optimization needs.
• Reduce database load, infrastructure cost, and user-facing latency.
• Proactively manage performance rather than reacting to production issues.
• Build a structured tuning workflow backed by data-driven insights.

When to use query recommendations

Use query recommendations as part of your regular tuning and performance management process:

• During initial onboarding of new applications.
• When migrating databases to new platforms or versions.
• After significant schema or application logic changes.
• When monitoring for performance regressions or capacity trends.
• As part of ongoing database health reviews — query patterns evolve even without code changes.
• During release validation — check whether new workload patterns have introduced tuning opportunities.

How query recommendations work

Hybrid Manager collects query diagnostics data, including:

• Query latency and execution patterns.

• Wait state analysis

• Workload trends and query frequency.

• The Query Diagnostics tab — where recommendations are shown inline with individual queries.

• The Recommendations tab — where all current recommendations for the selected database or cluster are listed.

You can move seamlessly from diagnosing a problematic query to understanding and applying the relevant recommendation.

Integrating query recommendations into your tuning process

You can incorporate Hybrid Manager recommendations into your broader tuning workflow:

• Daily or weekly reviews: Regularly check the Recommendations tab to identify new opportunities.
• Targeted tuning: Focus on high-latency or high-frequency queries flagged in Query Diagnostics. Start by applying high-impact recommendations (such as missing indexes).
• Trend analysis: Monitor how query performance evolves after applying recommendations. Hybrid Manager helps you track whether tuning efforts are effective.
• Cross-team collaboration: Developers, DBAs, and platform engineers can use the same recommendations to prioritize and coordinate tuning work.
• Optimization backlog: Treat unresolved recommendations as a backlog of tuning work to tackle during capacity planning cycles.

Workflow: From query to optimization

To use query recommendations:

1. Open the Query Diagnostics tab for your cluster.
2. Select a time range and review query performance.
3. Look for queries with associated recommendations — these are visually indicated in the query list.
4. Select the recommendation to open the Recommendations tab, with the relevant recommendation highlighted.
5. Review the recommendation details, including:

• The impact on the selected query.
• Any other queries that will benefit.
• Guidance on applying the optimization.

6. Apply the recommendation (for example, by creating an index or updating statistics).
7. Monitor the query over time to validate performance improvements.

Requirements for external clusters

To view query recommendations for external (non-Hybrid Manager-managed) clusters:

• The *Query Diagnostics extension must be enabled.
• The Query Advisor extension must be enabled.
• The Hybrid Manager beacon agent must be installed and running.

Without these components, Hybrid Manager cannot collect the necessary data to generate recommendations.

Example scenarios

Here are typical tuning scenarios where query recommendations provide value:

• Slow application pages: A frequently used query with high latency is identified in Query Diagnostics. A recommendation suggests adding an index on a filter column. Once applied, response times improve significantly.
• ETL job optimization: A batch process is running slower after recent data growth. Query Diagnostics shows an aggregate query performing poorly. A statistics refresh recommendation improves planner accuracy, restoring job performance.
• Capacity planning: Before scaling hardware, you review Recommendations to see whether tuning can alleviate bottlenecks without adding compute.
• Proactive tuning: During a routine tuning cycle, you review recommendations across critical databases and systematically apply improvements. As a result, you reduce overall database load by optimizing expensive queries.
• Release validation: After deploying new application features, you check whether new query patterns have surfaced tuning opportunities. Applying recommendations early prevents future performance issues.

Supported recommendation types

Hybrid Manager currently supports:

• Index recommendations — such as adding a missing index on a frequently used column.
• Statistics recommendations — such as refreshing outdated statistics to improve planner accuracy.

Additional recommendation types are planned for future releases.

Related information

• Query Diagnostics
• Recommendations tab