Skip to content

[Storage] Implement batch ingestion API for high-volume RDF datasets #615

Description

@remiceres

Summary

Provide an optimized batch ingestion API in core.next.storage to ingest high-volume RDF graphs without memory exhaustion or excessive garbage collection overhead.

Objectives and Technical Scope

1. Batch Ingestion Interface

• Design a high-throughput batch loading contract (addBatch(Iterable<Statement>) / stream consumer) in core.next.storage.api.
• Implement bulk index population that minimizes pointer re-allocations and tree rebalancing overhead.

2. Index Deferral and Rebuilding

• Support temporary index suspension or bulk insertion modes during massive loads, followed by deterministic index finalization.

Acceptance Criteria

[ ] Bulk loading of 100,000+ statements completes without memory exhaustion or GC stalls.
[ ] All index invariants (SPO, POS, OSP) are fully satisfied and consistent post-ingest.
[ ] Memory footprint per statement during batch loading is measurably optimized compared to one-by-one insertion.

Activity

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Metadata

Metadata

Assignees

Labels

featureNew feature or capabilityrefactoringInternal code restructuring and architecture cleanup

Type

No type

Projects

No projects

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions