toto+base+all - Analysis Report
Generated: 2026-02-12 10:53:30
Benchmark Run: 2026-01-21T03:41:05.573166
Description: TOTO Study: All tests with toto + baseline (without toto)
Total Executions: 1080
Summary Statistics
Execution Overview
| Metric | Value |
|---|---|
| Benchmark Name | toto+base+all |
| Search Method | exhaustive |
| Total Executions | 180 × 6 = 1080 |
| Total Runtime | 19.62h |
Metrics Overview
| Metric | Min | Max | Mean | Std Dev |
|---|---|---|---|---|
| bwMiB | 2189.6264 | 10898.9941 | 8699.8935 | 1226.6133 |
| totalTime | 1.5033 | 7.4826 | 1.9296 | 0.3619 |
Execution Details
| Metric | Value |
|---|---|
| IOPS Version | 3.5.0.dev (report: v3.5.1.dev) |
| Hostname | devel02.plafrim.cluster |
| Executor | slurm |
| Start Time | 2026-01-20 08:03:51 |
| End Time | 2026-01-21 03:41:05 |
| Avg Cores per Execution | 576.0 |
Search Space
| Variable | Type | Values |
|---|---|---|
| with_toto | bool | False, True |
| nodes | int | 16 |
| total_procs | int | 16, 64, 256 |
| requests_mb | int | 1, 8, 32 |
| volume_size_gb | int | 16 |
| analysis_period | int | 0, 250, 500, 1000, 2000 |
| spatiality | str | cont, random |
| filestrategy | str | file-per-proc, shared-file |
System Environment
Information collected from compute nodes during benchmark execution:
Hardware Summary
| Property | Value |
|---|---|
| Compute Nodes | 18 nodes (click to expand)bora001.plafrim.cluster, bora002.plafrim.cluster, bora009.plafrim.cluster, bora025.plafrim.cluster, bora005.plafrim.cluster, bora028.plafrim.cluster, bora026.plafrim.cluster, bora022.plafrim.cluster, bora003.plafrim.cluster, bora016.plafrim.cluster, bora012.plafrim.cluster, bora004.plafrim.cluster, bora010.plafrim.cluster, bora015.plafrim.cluster, bora011.plafrim.cluster, bora021.plafrim.cluster, bora006.plafrim.cluster, bora008.plafrim.cluster |
| CPU Model | Intel(R) Xeon(R) Gold 6240 CPU @ 2.60GHz |
| CPU Cores/Node | 36 |
| Memory/Node | 187.2 GB |
| Operating System | CentOS Linux 7 (Core) |
| Kernel | 3.10.0-957.el7.x86_64 |
Parallel Filesystems
| Type | Mount Point |
|---|---|
| beegfs | /beegfs |
Interconnect
InfiniBand Devices: hfi1_0
Best Configurations
Top 10 configurations for each metric (minimum 3 samples required):
Best for bwMiB
| Rank | with_toto | nodes | total_procs | requests_mb | volume_size_gb | analysis_period | bwMiB (mean) | Std Dev | Samples |
|---|---|---|---|---|---|---|---|---|---|
| 1 | True | 16 | 256 | 8 | 16 | 1000 | 9362.3334 | 1038.4482 | 24 |
| Command: ior -w -b 64mb -t 8mb -F -O summaryFile=/home/lgouveia/workdir/run_004/runs/exec_0155/repetition_001/summary_155_1.json -O summaryFormat=JSON -o /beegfs/lgouveia/ior_8/output.ior | |||||||||
| 2 | False | 16 | 256 | 8 | 16 | 0 | 9322.0555 | 1142.0977 | 24 |
| Command: ior -w -b 64mb -t 8mb -z -O summaryFile=/home/lgouveia/workdir/run_004/runs/exec_0031/repetition_001/summary_31_1.json -O summaryFormat=JSON -o /beegfs/lgouveia/ior_8/output.ior | |||||||||
| 3 | True | 16 | 64 | 32 | 16 | 500 | 9301.0326 | 961.0085 | 24 |
| Command: ior -w -b 256mb -t 32mb -z -F -O summaryFile=/home/lgouveia/workdir/run_004/runs/exec_0130/repetition_001/summary_130_1.json -O summaryFormat=JSON -o /beegfs/lgouveia/ior_8/output.ior | |||||||||
| 4 | True | 16 | 256 | 32 | 16 | 1000 | 9245.5759 | 1205.0700 | 24 |
| Command: ior -w -b 64mb -t 32mb -F -O summaryFile=/home/lgouveia/workdir/run_004/runs/exec_0171/repetition_001/summary_171_1.json -O summaryFormat=JSON -o /beegfs/lgouveia/ior_8/output.ior | |||||||||
| 5 | True | 16 | 256 | 8 | 16 | 250 | 9236.1845 | 1224.0726 | 24 |
| Command: ior -w -b 64mb -t 8mb -z -O summaryFile=/home/lgouveia/workdir/run_004/runs/exec_0157/repetition_001/summary_157_1.json -O summaryFormat=JSON -o /beegfs/lgouveia/ior_8/output.ior | |||||||||
| 6 | True | 16 | 64 | 8 | 16 | 2000 | 9218.0630 | 1018.8277 | 24 |
| Command: ior -w -b 256mb -t 8mb -z -O summaryFile=/home/lgouveia/workdir/run_004/runs/exec_0112/repetition_001/summary_112_1.json -O summaryFormat=JSON -o /beegfs/lgouveia/ior_8/output.ior | |||||||||
| 7 | True | 16 | 256 | 8 | 16 | 2000 | 9211.3275 | 1195.3005 | 24 |
| Command: ior -w -b 64mb -t 8mb -F -O summaryFile=/home/lgouveia/workdir/run_004/runs/exec_0156/repetition_001/summary_156_1.json -O summaryFormat=JSON -o /beegfs/lgouveia/ior_8/output.ior | |||||||||
| 8 | True | 16 | 64 | 32 | 16 | 250 | 9180.9936 | 977.6716 | 24 |
| Command: ior -w -b 256mb -t 32mb -F -O summaryFile=/home/lgouveia/workdir/run_004/runs/exec_0121/repetition_001/summary_121_1.json -O summaryFormat=JSON -o /beegfs/lgouveia/ior_8/output.ior | |||||||||
| 9 | True | 16 | 64 | 32 | 16 | 2000 | 9172.9989 | 1170.3237 | 24 |
| Command: ior -w -b 256mb -t 32mb -F -O summaryFile=/home/lgouveia/workdir/run_004/runs/exec_0124/repetition_001/summary_124_1.json -O summaryFormat=JSON -o /beegfs/lgouveia/ior_8/output.ior | |||||||||
| 10 | True | 16 | 256 | 32 | 16 | 500 | 9160.9235 | 1100.7386 | 24 |
| Command: ior -w -b 64mb -t 32mb -z -O summaryFile=/home/lgouveia/workdir/run_004/runs/exec_0174/repetition_001/summary_174_1.json -O summaryFormat=JSON -o /beegfs/lgouveia/ior_8/output.ior | |||||||||
Best for totalTime
| Rank | with_toto | nodes | total_procs | requests_mb | volume_size_gb | analysis_period | totalTime (mean) | Std Dev | Samples |
|---|---|---|---|---|---|---|---|---|---|
| 1 | True | 16 | 256 | 8 | 16 | 1000 | 1.7711 | 0.1994 | 24 |
| Command: ior -w -b 64mb -t 8mb -F -O summaryFile=/home/lgouveia/workdir/run_004/runs/exec_0155/repetition_001/summary_155_1.json -O summaryFormat=JSON -o /beegfs/lgouveia/ior_8/output.ior | |||||||||
| 2 | True | 16 | 64 | 32 | 16 | 500 | 1.7801 | 0.1878 | 24 |
| Command: ior -w -b 256mb -t 32mb -z -F -O summaryFile=/home/lgouveia/workdir/run_004/runs/exec_0130/repetition_001/summary_130_1.json -O summaryFormat=JSON -o /beegfs/lgouveia/ior_8/output.ior | |||||||||
| 3 | False | 16 | 256 | 8 | 16 | 0 | 1.7834 | 0.2208 | 24 |
| Command: ior -w -b 64mb -t 8mb -z -O summaryFile=/home/lgouveia/workdir/run_004/runs/exec_0031/repetition_001/summary_31_1.json -O summaryFormat=JSON -o /beegfs/lgouveia/ior_8/output.ior | |||||||||
| 4 | True | 16 | 64 | 8 | 16 | 2000 | 1.7988 | 0.2029 | 24 |
| Command: ior -w -b 256mb -t 8mb -z -O summaryFile=/home/lgouveia/workdir/run_004/runs/exec_0112/repetition_001/summary_112_1.json -O summaryFormat=JSON -o /beegfs/lgouveia/ior_8/output.ior | |||||||||
| 5 | True | 16 | 256 | 32 | 16 | 1000 | 1.8030 | 0.2483 | 24 |
| Command: ior -w -b 64mb -t 32mb -F -O summaryFile=/home/lgouveia/workdir/run_004/runs/exec_0171/repetition_001/summary_171_1.json -O summaryFormat=JSON -o /beegfs/lgouveia/ior_8/output.ior | |||||||||
| 6 | True | 16 | 64 | 32 | 16 | 250 | 1.8047 | 0.1981 | 24 |
| Command: ior -w -b 256mb -t 32mb -F -O summaryFile=/home/lgouveia/workdir/run_004/runs/exec_0121/repetition_001/summary_121_1.json -O summaryFormat=JSON -o /beegfs/lgouveia/ior_8/output.ior | |||||||||
| 7 | True | 16 | 256 | 8 | 16 | 250 | 1.8065 | 0.2589 | 24 |
| Command: ior -w -b 64mb -t 8mb -z -O summaryFile=/home/lgouveia/workdir/run_004/runs/exec_0157/repetition_001/summary_157_1.json -O summaryFormat=JSON -o /beegfs/lgouveia/ior_8/output.ior | |||||||||
| 8 | True | 16 | 256 | 8 | 16 | 2000 | 1.8098 | 0.2527 | 24 |
| Command: ior -w -b 64mb -t 8mb -F -O summaryFile=/home/lgouveia/workdir/run_004/runs/exec_0156/repetition_001/summary_156_1.json -O summaryFormat=JSON -o /beegfs/lgouveia/ior_8/output.ior | |||||||||
| 9 | True | 16 | 256 | 32 | 16 | 500 | 1.8147 | 0.2291 | 24 |
| Command: ior -w -b 64mb -t 32mb -z -O summaryFile=/home/lgouveia/workdir/run_004/runs/exec_0174/repetition_001/summary_174_1.json -O summaryFormat=JSON -o /beegfs/lgouveia/ior_8/output.ior | |||||||||
| 10 | False | 16 | 64 | 32 | 16 | 0 | 1.8154 | 0.1959 | 24 |
| Command: ior -w -b 256mb -t 32mb -z -F -O summaryFile=/home/lgouveia/workdir/run_004/runs/exec_0024/repetition_001/summary_24_1.json -O summaryFormat=JSON -o /beegfs/lgouveia/ior_8/output.ior | |||||||||
Exhaustive Search Exploration
This section shows the exploration behavior of the exhaustive search. All parameter combinations are tested systematically. Running maximum and minimum lines show the range of values discovered over execution order.
bwMiB Evolution
Shows metric values in execution order. Running max and min lines help visualize the range of values discovered over time.
totalTime Evolution
Shows metric values in execution order. Running max and min lines help visualize the range of values discovered over time.
Variable Analysis & Relationships
Analysis of how variables relate to performance metrics.
Variable Impact on Performance
Statistical Method: Variance decomposition analysis
The impact score quantifies the proportion of total variance in the performance metric that is explained by each variable. For each variable, the metric values are grouped by the variable's levels, and the between-group variance is computed. The impact score is calculated as:
Impact = σ²_between / σ²_total
where σ²_between is the variance of group means (weighted by group size) and σ²_total is the overall variance of the metric. This measure is analogous to the R² coefficient in regression or the eta-squared (η²) effect size in ANOVA. Values range from 0 (no effect) to 1 (variable fully determines the metric).
Custom Metric Plots
User-defined plots from the reporting configuration.