Performance Optimization Flashcards

In IICS, what does the 'session recovery' feature do when a long-running mapping fails midway, and how does it affect performance design?

Answer: It resumes processing from the last recovery checkpoint, requiring target tables to support idempotent or incremental loads

Session recovery restarts from the last saved checkpoint, but the target must handle partial reloads gracefully to avoid duplicate data.

Which IICS optimizer hint tells the Secure Agent to use a pipeline partition strategy based on key range rather than hash?

Answer: Key range partitioning

Key range partitioning divides rows into segments based on value ranges of a chosen key column, which can improve locality for sorted data sets.

An IICS mapping that calls an external REST API for each row is running slowly. What architectural change delivers the greatest throughput improvement?

Answer: Batch multiple rows into a single API request using a payload builder expression

Batching multiple rows into one HTTP call drastically reduces network round-trips and API overhead compared to one call per row.

In Informatica Cloud CDI, what is the function of the 'Early Selection' optimization?

Answer: It pushes a filter as close to the source as possible to reduce rows flowing through the pipeline

Early Selection moves filter conditions upstream so fewer rows enter subsequent transformations, reducing total processing work.

When connecting IICS to an on-premises Oracle database over a WAN with high latency, which Secure Agent tuning reduces the impact of network round-trips?

Answer: Increasing the JDBC fetch size (array fetch size) to retrieve more rows per round-trip

A larger array fetch size causes JDBC to retrieve hundreds of rows per network call, amortizing WAN latency across many records.

Which IICS Advanced Data Quality option should be disabled on a non-profiling run to improve mapping throughput?

Answer: Data profiling / profile statistics collection

Data profiling collects statistics on every row processed; disabling it on production runs removes that per-row overhead.

In an IICS mapping with multiple source tables joined via a Joiner, what is the impact of placing a Filter transformation before the Joiner rather than after it?

Answer: It reduces the number of rows that enter the Joiner, shrinking the join dataset and improving speed

Filtering rows before the Joiner means fewer rows need to be matched and cached, directly reducing join execution time and memory use.