Performance Improvements: Memory Leak Fixes, Label Indexing, and Optimized Edge & Buffer Handling

[ParameswaranSajeenthiran]

Description

This PR introduces several key performance and memory optimizations to improve system efficiency and scalability:

• Memory Leak Fixes:

  • Identified and removed memory leaks in critical components to ensure stable long-term operation and prevent unnecessary memory consumption.

• Label Indexing:

  • Added efficient label indexing for faster node/edge lookups and reduced query latency.

• Optimized Shared Buffer Usage:

  • Removed redundant lock waits in the single-threaded use of the shared buffer by introducing partition-wise producer-consumer handling.
  • This change minimizes thread contention and improves throughput for each partition.

• Edge Storage Optimization:

  • Removed redundant bidirectional edge cut storage, reducing both memory usage and I/O overhead without affecting query correctness.

---

Impact

• Performance: Reduced lock contention and improved buffer throughput.
• Memory: Lower memory footprint due to elimination of leaks and unnecessary edge storage.
• Scalability: Faster query execution with label indexing and optimized data structures.

---

Testing

• Verified memory stability through long-running stress tests.
• Benchmarked query performance showing reduced latency after label indexing.
• Validated correctness of producer-consumer operations and graph queries after removing redundant storage.

---

---

Related Issues

[sonarqubecloud]

Quality Gate failed

Failed conditions
26 Security Hotspots
15.9% Duplication on New Code (required ≤ 3%)

See analysis details on SonarQube Cloud