Results for “fnd_conc_stat_g_p_summary”

AI-generated from documented ETRM metadata — verify critical details on the linked pages.

Overview

FND_CONC_STAT_G_P_SUMMARY is an APPLSYS-owned table within the FND – Application Object Library product of Oracle E-Business Suite, valid in both 12.1.1 and 12.2.2. The table stores aggregated runtime statistics for concurrent programs, keyed by a statistic identifier, a grouping code, and the period over which the statistics were collected. It is a sibling of the well-known FND_CONC_STAT_SUMMARY and FND_CONC_STAT_G_SUMMARY tables; the "_P_" designation indicates that this variant segregates values by PROCESS_TYPE (for example, distinguishing manager versus worker or different concurrent processing tiers), producing a per-process-type rollup. The table answers the practical question of how a given concurrent program, request set, or group performed over a defined collection window, and it forms part of the foundation for concurrent manager tuning and the Concurrent: Summarize Concurrent Program Statistics and related concurrent programs.

Under the heuristic Data Vault classification supplied in the metadata, the object is designated standalone — meaning it is not modeled as a hub, link, or satellite in the source FK structure, but behaves as an independent aggregate/fact table. In a Data Vault modeling suggestion, its natural-key tuple (STATISTIC#, GROUP_CODE, COLLECT_PERIOD, and implicitly PROCESS_TYPE) would serve as the hub key, while the measured values would load into a satellite or a fact-style structure.

Key Information Stored

The table contains 12 documented columns. The most consequential are:

• STATISTIC# – Identifier of the statistic being captured (for example elapsed time, CPU time, or DB time). Part of the composite primary key.
• GROUP_CODE – Code identifying the grouping dimension against which statistics are aggregated. Part of the composite primary key.
• COLLECT_PERIOD – The collection window label over which the aggregate was computed. Part of the composite primary key.
• COLLECT_DATE – Timestamp at which the row was generated or refreshed.
• PROCESS_TYPE – The process-type discriminator that distinguishes this summary from the generic G summary table.
• MIN, MAX, MEAN – Minimum, maximum, and mean observed values across the sampled requests.
• N – Sample size (count of observations) underlying the aggregate.
• ROLLUP_VAR_N_1, ROLLUP_VAR_N_2, ROLLUP_VAR_C_1 – Numeric and character rollup/variance accumulators supporting higher-level aggregation of the collected statistics.

The primary key is FND_CONC_STAT_G_P_SUMMARY_PK over (STATISTIC#, GROUP_CODE, COLLECT_PERIOD). A unique index, FND_CONC_STAT_G_P_SUMMARY_U1, covers the identical column list and serves as the business-key candidate. No FK relationships are documented, consistent with the standalone classification.

Common Use Cases and Queries

Typical uses include concurrent processing performance reporting, capacity planning, and diagnosing skew between process types. A representative query returns the mean and maximum for a chosen statistic over recent collection periods:

• SELECT collect_period, process_type, min, max, mean, n FROM fnd_conc_stat_g_p_summary WHERE statistic# = :stat AND group_code = :grp ORDER BY collect_period DESC;
• SELECT process_type, AVG(mean), SUM(n) FROM fnd_conc_stat_g_p_summary WHERE collect_period BETWEEN :from AND :to GROUP BY process_type;

These patterns support historical trending, outlier detection (comparing MAX to MEAN), and confidence assessment via N. Because rows are keyed by period, they lend themselves to time-series BI extracts and to comparison of 12.1.1 versus 12.2.2 concurrent manager behavior.

Related Objects

The table participates in the family of concurrent statistics objects, though the metadata documents no explicit FK joins. Practically, it is queried alongside:

• FND_CONC_STAT_SUMMARY – Overall (non-grouped) concurrent statistics.
• FND_CONC_STAT_G_SUMMARY – Grouped concurrent statistics without the process-type split.
• FND_CONCURRENT_PROGRAMS – Definition of the concurrent programs whose execution generates the statistics.
• FND_CONCURRENT_REQUESTS – Individual request rows underlying the aggregates.
• FND_CONCURRENT_PROCESSES – The running manager/worker processes reflected in PROCESS_TYPE.
• FND_CP_SERVER_TYPES and FND_CONCURRENT_QUEUES – Concurrent manager and specialization configuration referenced by group codes.

Because the object is classified as standalone and no foreign keys are documented, joins to these tables should be constructed explicitly on STATISTIC#, GROUP_CODE, and PROCESS_TYPE rather than relying on enforced referential constraints.