Search Results collect_date

Overview

FND_CONC_STAT_SUMMARY is a table in the APPLSYS schema owned by the Application Object Library (FND) product within Oracle E-Business Suite 12.1.1 and 12.2.2. It stores aggregated runtime statistics for concurrent programs executed by the Concurrent Manager. Each row captures distributional metrics — count, minimum, maximum, mean, and standard deviation — for a single concurrent program's processing, segmented by process type and by whether the observation was collected on a daily basis. The table underpins the "Concurrent Program Statistics" functionality used by system administrators to evaluate performance and identify programs that consistently consume disproportionate processing time.

The ETRM metadata labels the object as "Retrofitted," indicating it is a migrated or re-registered object rather than a natively recorded ETRM entity; the physical schema is nonetheless documented with eleven columns. Heuristic Data Vault classification mined from the foreign-key structure identifies this table as standalone — it holds no inbound or outbound foreign-key relationships, and is therefore modeled as an independent table rather than a hub, link, or satellite. In Data Vault terms, no integration hub or linking construct is suggested; the statistics are purely descriptive and self-contained.

Key Information Stored

The primary key is FND_CONC_STAT_SUMMARY_PK, a composite key defined over five columns: STATISTIC#, PROGRAM_NAME, PROCESS_TYPE, DAILY, and COLLECT_DATE. Because the primary key is composite and drawn from business attributes rather than a generated surrogate identifier, PROGRAM_NAME and COLLECT_DATE act as the principal business-key candidates. The most significant columns are:

• STATISTIC# — Numeric identifier distinguishing individual statistic records within the composite key.
• PROGRAM_NAME — The concurrent program whose execution is being summarized; the primary business identifier.
• PROCESS_TYPE — The process or manager type associated with the run, allowing separation of statistics by execution category.
• DAILY — Flag or value indicating whether the statistic is accumulated daily rather than over a cumulative window.
• COLLECT_DATE — The date on which the statistics were collected; drives time-based trending and reporting.
• N — The number of observations (sample count) contributing to the summarized values.
• MINVAL — The minimum observed value in the collection window.
• MAXVAL — The maximum observed value in the collection window.
• MEAN — The arithmetic mean of the collected observations.
• SIGMA — The standard deviation, expressing variability across the observations.
• SS — The sum of squares, an aggregate used in variance and statistical computation.

Common Use Cases and Queries

The table supports concurrent processing performance analysis, capacity planning, and identification of runaway or long-running programs. A typical query retrieves mean and maximum runtimes by program for a collection period:

SELECT program_name, process_type, collect_date,
       n, minval, maxval, mean, sigma
FROM   applsys.fnd_conc_stat_summary
WHERE  collect_date >= :start_date
AND    daily = 'Y'
ORDER BY mean DESC;

Report consumers use MINVAL, MAXVAL, MEAN, and SIGMA to build control charts and detect programs whose execution deviates from historical norms. Administrators compare N across COLLECT_DATE values to gauge sampling reliability, discarding rows with insufficient observations. Trend reports join successive COLLECT_DATE snapshots to show whether a program's mean runtime is drifting upward, informing tuning or scheduling changes. Statistics gathered here also inform workload balancing decisions when concurrent managers are reassigned.

Related Objects

The mined relationship data classifies FND_CONC_STAT_SUMMARY as standalone, so no declarative foreign keys are documented. Logical relationships nonetheless exist with the surrounding concurrent processing infrastructure. PROGRAM_NAME corresponds to CONCURRENT_PROGRAMS.CONCURRENT_PROGRAM_NAME (FND_CONCURRENT_PROGRAMS), while PROCESS_TYPE aligns with the manager types recorded in FND_CONCURRENT_QUEUES. The statistic definitions referenced by STATISTIC# are conceptually tied to FND_CONC_STAT_DEFN and its associated FND_CONC_STAT_DEFN_TL translations. Execution-level source data that feeds these summaries resides in FND_CONCURRENT_REQUESTS and FND_CONCURRENT_PROGRAMS, with FND_CONCURRENT_WORKER and FND_CONCURRENT_QUEUE_CONTENT describing the managers and their assignments. Administrators typically join FND_CONCURRENT_PROGRAMS and FND_CONCURRENT_REQUESTS to correlate summarized statistics with the actual request records from which they were derived.