benchmark_config_template.toml

# CHIMERA Benchmark Configuration Template
# Hardware Profiling and Dataset Transparency

# This template provides a structured format for documenting all
# benchmark configurations, hardware specifications, and dataset
# parameters to ensure full reproducibility per ACM standards.

# ============================================================
# BENCHMARK METADATA
# ============================================================

[metadata]
benchmark_suite = "CHIMERA"
version = "1.0.0"
timestamp = "2026-01-19T11:14:35Z"  # ISO 8601 format
run_id = "unique-run-identifier"
operator = "researcher-name"
organization = "institution-name"

# ============================================================
# HARDWARE PROFILE
# ============================================================

[hardware]
# Automatically collected via HardwareProfile.collect()
# Can be manually specified for validation

# System Information
hostname = "benchmark-system-01"
platform = "Linux-6.1.0-x86_64"  # OS-Version-Architecture
architecture = "x86_64"

# CPU Specifications
[hardware.cpu]
model = "Intel(R) Core(TM) i9-13900K"
vendor = "GenuineIntel"
cores_physical = 24
cores_logical = 32
frequency_ghz = 5.8
frequency_base_ghz = 3.0
cache_l1_kb = 80
cache_l2_kb = 2048
cache_l3_kb = 36864
features = ["AVX2", "AVX512", "FMA3"]

# Memory Specifications
[hardware.memory]
total_gb = 64.0
available_gb = 60.5
type = "DDR5"
speed_mhz = 5600
channels = 2
swap_total_gb = 16.0

# Storage Specifications
[hardware.storage]
type = "NVMe"
model = "Samsung 990 PRO"
capacity_gb = 2048
interface = "PCIe 4.0 x4"
sequential_read_mb_s = 7450
sequential_write_mb_s = 6900
random_read_iops = 1400000
random_write_iops = 1550000
filesystem = "ext4"
mount_options = "noatime,data=ordered"

# GPU Specifications (if applicable)
[[hardware.gpu]]
index = 0
name = "NVIDIA GeForce RTX 4090"
memory_gb = 24
cuda_version = "12.3"
compute_capability = "8.9"
driver_version = "545.29.06"
pcie_generation = 4
pcie_lanes = 16

[[hardware.gpu]]
index = 1
name = "NVIDIA GeForce RTX 4090"
memory_gb = 24
cuda_version = "12.3"
compute_capability = "8.9"
driver_version = "545.29.06"
pcie_generation = 4
pcie_lanes = 16

# Network Configuration
[hardware.network]
interfaces = ["eth0", "lo"]
bandwidth_gbps = 10.0
latency_us = 0.1
type = "Ethernet"

# ============================================================
# SOFTWARE ENVIRONMENT
# ============================================================

[software]
os_distribution = "Ubuntu 22.04.3 LTS"
kernel_version = "6.1.0-1027-aws"
compiler = "GCC 11.4.0"
compiler_flags = "-O3 -march=native -mtune=native"
build_type = "Release"

# Database Version
database_name = "System Under Test"
database_version = "1.0.0"
database_commit = "abc123def456"

# Dependencies
[software.dependencies]
rocksdb = "8.9.1"
llama_cpp = "b1696"
cuda_toolkit = "12.3"
cudnn = "8.9.7"
python = "3.11.6"
numpy = "1.26.2"
scipy = "1.11.4"

# ============================================================
# BENCHMARK CONFIGURATION
# ============================================================

[benchmark]
category = "transactional"  # transactional, analytical, vector, graph, llm, rag, system, multi-model
test_name = "YCSB_WorkloadA"
standard_reference = "YCSB"  # Reference benchmark standard

# Execution Parameters
repetitions = 10
iterations_per_run = 100
warmup_runs = 5
timeout_seconds = 300
random_seed = 42
deterministic = true

# Threading Configuration
thread_count = 8
thread_affinity = true
cpu_pinning = [0, 1, 2, 3, 4, 5, 6, 7]

# Memory Configuration
batch_size = 32
buffer_size_mb = 256
cache_size_mb = 1024

# ============================================================
# DATASET SPECIFICATION
# ============================================================

[dataset]
name = "ycsb_workload_a"
type = "synthetic"  # synthetic or real-world
size_mb = 1024
record_count = 1000000
generation_method = "ycsb_generator"
generation_script = "scripts/generate_ycsb_data.sh"
random_seed = 42
checksum_sha256 = "abcdef1234567890abcdef1234567890abcdef1234567890abcdef1234567890"
generation_timestamp = "2026-01-19T10:00:00Z"

# Dataset Parameters
[dataset.parameters]
field_count = 10
field_length = 100
record_size_bytes = 1000
key_distribution = "zipfian"
zipfian_constant = 0.99
read_proportion = 0.50
update_proportion = 0.50
insert_proportion = 0.00
scan_proportion = 0.00

# Data Characteristics
[dataset.characteristics]
unique_keys = 1000000
total_records = 1000000
size_on_disk_mb = 1024
compression = "none"
data_format = "key-value"

# ============================================================
# DATABASE CONFIGURATION
# ============================================================

[database_config]
cache_size_mb = 1024
write_buffer_size_mb = 256
max_write_buffer_number = 4
max_background_jobs = 8
compression_type = "zstd"
compression_level = 3
block_size_kb = 16
bloom_filter_bits_per_key = 10
enable_statistics = true
enable_profiling = true

# RocksDB Specific Options
[database_config.rocksdb]
max_open_files = -1
allow_mmap_reads = false
allow_mmap_writes = false
use_direct_reads = true
use_direct_io_for_flush_and_compaction = true
compaction_style = "level"
level0_file_num_compaction_trigger = 4
level0_slowdown_writes_trigger = 20
level0_stop_writes_trigger = 36

# ============================================================
# STATISTICAL CONFIGURATION
# ============================================================

[statistics]
# Outlier Detection
remove_outliers = true
outlier_method = "iqr"  # iqr, zscore, modified_zscore
iqr_multiplier = 1.5
zscore_threshold = 3.0

# Confidence Intervals
confidence_level = 0.95  # 95% CI
alternative_ci = 0.99    # 99% CI also reported

# Significance Testing
significance_alpha = 0.05  # α = 0.05 (95% confidence)
test_method = "t-test"     # t-test, mann-whitney, anova
two_tailed = true

# Effect Size
effect_size_method = "cohens_d"  # cohens_d, hedges_g
interpret_effect_size = true

# Power Analysis
target_power = 0.80
minimum_detectable_effect = 0.5  # Cohen's d = 0.5 (medium effect)

# ============================================================
# REPRODUCIBILITY CHECKLIST
# ============================================================

[reproducibility]
# ACM Artifact Badging Compliance
artifacts_available = true
artifacts_functional = true
results_reproducible = true

# Documentation
methodology_documented = true
hardware_documented = true
software_documented = true
dataset_documented = true
parameters_documented = true

# Version Control
code_repository = "https://github.com/your-org/your-system"
commit_hash = "abc123def456"
tag = "v1.0.0"

# Data Availability
data_generation_scripts = true
raw_results_available = true
analysis_scripts_available = true

# Environment
containerized = true
container_image = "your-org/system-benchmark:1.0.0"
container_registry = "docker.io"

# ============================================================
# PERFORMANCE TARGETS (Optional)
# ============================================================

[targets]
# Define expected performance targets for validation
throughput_ops_per_sec = 100000
latency_p50_ms = 2.0
latency_p95_ms = 5.0
latency_p99_ms = 10.0
memory_usage_mb = 1024
cpu_utilization_percent = 80.0

# ============================================================
# NOTES AND OBSERVATIONS
# ============================================================

[notes]
# Free-form notes about the benchmark run
pre_test_observations = """
System was in idle state before benchmark.
All non-essential services stopped.
CPU governor set to 'performance'.
"""

post_test_observations = """
No thermal throttling observed.
Memory usage remained stable throughout test.
No I/O errors logged.
"""

known_issues = """
None
"""

# ============================================================
# CONTACT INFORMATION
# ============================================================

[contact]
researcher = "Researcher Name"
email = "researcher@institution.edu"
institution = "Research Institution"
project = "CHIMERA Benchmark Suite"
funding = "Grant Number (if applicable)"

# ============================================================
# VALIDATION SIGNATURE
# ============================================================

[validation]
# For multi-party validation and review
reviewed_by = ["reviewer1@domain.com", "reviewer2@domain.com"]
review_date = "2026-01-20"
validation_notes = """
Configuration reviewed and approved.
Hardware specifications verified.
Dataset integrity confirmed.
"""

signature_hash = "signature_placeholder"