Loading...

XML

Word

Printable

JSON

Scale test phase T8a (~5 days). Large dataset load and memory verification.

Test coverage:

Load tables with 10M, 100M, 1B rows using bulk insert flush
Measure throughput (rows/sec) vs regular cursor insert at each scale
Verify O(1) memory: RSS during bulk insert must not grow proportionally to row count
Measure and report: flush latency, number of flushes triggered, bytes written per flush
Baseline comparison: bulk insert flush vs existing WiredTiger bulk loader

Depends on: WT-17215 (T4)
Can run in parallel with: T8b (WT-17222)

is related to

WT-17215 Bulk insert flush — implement production __wt_bulk_flush() orchestrator

WT-17218 Bulk insert flush — cppsuite stress test: concurrent bulk loads with background checkpoint

WT-17219 Bulk insert flush — cppsuite stress test: crash injection during flush

WT-17220 Bulk insert flush — cppsuite stress test: memory pressure

WT-17222 Bulk insert flush — performance benchmarking: throughput and flush overhead