Skip to main content

ASH-like workload visibility. Without querying ASH.

Oracle Active Session History is powerful. It is also tied to the Oracle Diagnostics Pack and is normally explored through Oracle Enterprise Manager.

Harry the Performance Scraper takes a different approach: it periodically samples live session data, stores the observations in PostgreSQL, and makes them available through Grafana. It does not query V$ACTIVE_SESSION_HISTORY or DBA_HIST_ACTIVE_SESS_HISTORY.

The following comparisons show Harry and the ASH Analytics view in Oracle Enterprise Manager 24ai for the same databases and time ranges.

Harry compared with Oracle Enterprise Manager ASH Analytics: sustained database workload

The visual result speaks for itself. Harry identifies the same:

  • workload start and end;
  • peaks and changes in database activity;
  • dominant wait classes;
  • top SQL identifiers;
  • top sessions responsible for the activity.

The charts are not expected to be pixel-for-pixel identical. Harry and ASH use different sampling intervals and aggregation methods. What matters during an investigation is that the relevant workload pattern—and the sessions and SQL behind it—is clearly visible in both.

Harry compared with Oracle Enterprise Manager ASH Analytics: intermittent activity peaks

Same workload, clearer workflow​

Returning to Enterprise Manager after working with Grafana makes the usability difference difficult to ignore.

Grafana provides fast time-range selection, zooming, refresh controls, consistent dashboard variables and several correlated views on the same screen. Harry uses that experience to place the activity timeline, Top SQL, Top Sessions, SQL activity over time and the underlying samples together.

Instead of navigating through multiple OEM screens, the investigation starts with a single question:

What happened in the database, when did it happen, and who was responsible?

Harry compared with Oracle Enterprise Manager ASH Analytics: several distinct workload periods

Useful visibility without licensed ASH data​

ASH remains a valuable Oracle diagnostic feature. Harry is not attempting to pretend otherwise.

Its goal is to provide practical historical workload visibility in environments where querying ASH is unavailable, undesirable, or difficult to justify. These side-by-side captures show that external session sampling is not merely a simplified activity counter: it can preserve the workload shape and retain the evidence needed to identify the main SQL and sessions involved.

Harry compared with Oracle Enterprise Manager ASH Analytics: matching workload spikes and dominant SQL

The result is a lightweight and accessible starting point for Oracle performance analysis:

the workload pattern, the responsible SQL, and the responsible sessions—all visible without querying ASH.