DB2 Performance Evolution: Comprehensive Analysis of Versions 10.2.4, 10.1 FP1, and 10.1 GA

​This comprehensive analysis compares the performance evolution across three HWA (Workload Automation) releases using DB2 as the database backend, demonstrating the progression from 10.1 GA through 10.1 FP1 to the current 10.2.4 release.

​Executive Summary
- HWA 10.2.4 with DB2 delivers exceptional performance improvements over previous versions
- Significant reductions in scheduling and mirroring delays during extended loads
- Consistent throughput capabilities maintained across all versions
- Mixed results in critical operations with some areas showing improvement potential

---

Test Environment Configuration
All tests maintained consistent workload parameters for fair comparison:
Common Configuration:
- Workload: 600K+ jobs across 100+ workstations and 43K+ job streams
- Peak Load: ~2,650 jobs/min during burst periods
- Extended Load: 1,200 jobs/min sustained for 5 hours
- Database: DB2 backend across all versions
Version-Specific Infrastructure:
- 10.2.4: VMware 6, RHEL 9.4, DB2 11.5.9, WebSphere Liberty 25.0.0.3
- 10.1 FP1: VMware ESXi 7.0.3, RHEL 8.3, DB2 11.5.8, WLP 22.0.0.12/13
- 10.1 GA: VMware ESXi 7.0.3, RHEL 8.3, DB2 11.5.8, WLP 22.0.0.12

Performance Metrics Comparison
🚀 Peak Load Performance (10-Minute Burst)
​

​⚡ Extended Load Performance (5-Hour Sustained)

​🔧 Critical Operations Performance

Performance Evolution Analysis

Outstanding Improvements in 10.2.4
🎯 Schedule Delay Optimization: The most significant improvement across all scenarios:
Peak Load:
- 38% better than 10.1 FP1 (34s vs 55s)
- 29% better than 10.1 GA (34s vs 48s)
Extended Load:
- 60% better than 10.1 FP1 (17s vs 43s)
- 53% better than 10.1 GA (17s vs 36s)
🏆 Resource Utilization: Significant improvements in DB2 performance:
- DB2 CPU: 11% reduction during peak loads
- DB2 Disk: 33% reduction in disk utilization
- Consistent throughput maintained across all versions
⚡Extended Load Excellence: 10.2.4 shows remarkable stability during sustained workloads with both schedule and mirroring delays remaining well under 20 seconds.

Areas Requiring Attention
📈 Peak Load Mirroring: 10.2.4 shows higher mirroring delays during peak loads (61s vs 37s in FP1), though still within acceptable limits.
🔧Critical Operations: Both MAKEPLAN and CHECKSYNC operations maintain consistent performance across all versions:
- MAKEPLAN: Stable 4 minutes across all versions
- CHECKSYNC: Consistent 7 minutes across all versions

Version Progression Insights
10.1 GA → 10.1 FP1: Mixed results with some regressions in delay metrics but improvements in cleanup operations.
10.1 FP1 → 10.2.4: Dramatic improvements in core performance metrics, particularly schedule delays, with some trade-offs in critical operations timing.

---

Recommendations

For Production Deployment
✅ Highly Recommended: HWA 10.2.4 with DB2 is ideal for:
- High-volume production environments requiring consistent low-latency scheduling
- Extended load scenarios where sustained performance is critical
- DB2-optimized environments benefiting from improved resource utilization

Implementation Considerations
⚠️ Planning Required:
- Monitor peak load mirroring performance in high-burst scenarios

Performance Optimization
🔧 Immediate Actions:
- Monitor DB2 resource utilization to leverage the improved efficiency
- Consider peak load scheduling patterns to minimize mirroring delay impact

---

Conclusion
HWA 10.2.4 with DB2 represents a major performance breakthrough, delivering exceptional improvements in the core scheduling and mirroring operations that define day-to-day workload automation performance. The 60% improvement in extended load schedule delays and 33% reduction in DB2 disk utilization make it the clear choice for enterprise deployments.
The consistent performance of critical operations (MAKEPLAN and CHECKSYNC) combined with the overall performance gains in production workload handling make version 10.2.4 the optimal choice for organizations prioritizing consistent, high-performance workload automation with DB2.
The evolution demonstrates a clear focus on optimizing the most frequent operations (job scheduling and mirroring) while maintaining stable throughput capabilities across all versions

---

Analysis based on comprehensive testing with 600K+ job workloads on DB2 11.5.x across enterprise VMware infrastructure

---

​Alessandro Tomasi
Senior Technical Lead
Alessandro is a Software Engineer. He has an academic background in Computer Engineering, in particular in technology customer support and automation testing

---

Marco Gangemi
​Software Engineer
Marco comes from an academic background in Physics.
He joined HCL in 2018.Since then, he has specialized in DevSecOps activities.
​He’s recently started on Performance tests.

---