Based on benchmarking 118,449 real gRPC protobuf messages (median 173 bytes):
| Dictionary Size | Compression Ratio | Bandwidth Saved | Performance |
|---|
Oddur Magnusson oddur
oddur
/ grpc-compression-analysis.md
Last active
October 12, 2025 20:53
gRPC Message Compression Analysis with Zstandard - Optimal compression level for production servers
oddur
/ simtime-multi-instance-architecture.md
Last active
October 28, 2025 15:11
SimTime Service Multi-Instance Architecture Plan - Leader-Follower approach for horizontal scaling
This document outlines a solution to enable the SimTime service to run safely across multiple instances, eliminating the current single point of failure while maintaining time consistency across the distributed system.
oddur
/ timestamp-optimization-analysis.md
Last active
November 7, 2025 10:26
NemoPostgresAdapter: Persistence Efficiency Optimization - Comprehensive analysis of batching inefficiencies and proposed accumulation strategy
The NemoPostgresAdapter persistence layer is exhibiting multiple inefficiency symptoms that indicate a fundamental batching problem. These issues were identified through:
- Performance benchmarking of the persistence layer under various load conditions
- PostgreSQL monitoring using
pg_stat_statements(query performance statistics) - Database activity analysis using
pg_stat_activity(connection and transaction patterns) - CPU profiling during idle and load scenarios