SoroushVahidi/llm-serving-scheduler-baselines

LLM-Serving Scheduler Baselines: Simulation Performance Outcomes for 13 Scheduling Policies

This is a comprehensive, text-free, highly structured simulation results dataset for large language model (LLM) serving schedulers. It contains 36,975 detailed policy-level outcome records generated across 41 distinct workload regimes and 2,175 traffic seeds. The dataset compares 12 highly optimized third-party baseline schedulers against APT-Serve (a transition-aware preemption-capable scheduler), complete with enriched nullable cache metrics (apt_stats).

This local staging release is version v1.

What One Row Means

In per_policy_results, one row represents the final end-to-end simulated performance metrics of one scheduler policy executing on one traffic seed under one specific workload regime/envelope.

Configurations

• per_policy_results: 36,975 rows. The flat main results table.

Workload Regimes

The dataset represents evaluations across 41 diverse, highly challenging workload envelopes designed to Stress-Test scheduling architectures. These are grouped into:

• Baseline Pressure: pressure_low_baseline, pressure_medium_baseline, pressure_high_baseline, pressure_near_capacity_baseline, pressure_sustained_overload_baseline.
• Arrival Patterns: arrival_steady, arrival_burst, arrival_clustered_burst, arrival_spike, arrival_alternating.
• Sequence Length Shifts: length_homogeneous, length_bimodal, length_prompt_heavy, length_decode_heavy, length_heavy_tail_prompt, length_heavy_tail_output.
• Cache Pressure and Thrashing Risk: cacheuse_recompute_avoidance, cacheuse_kv_to_hidden_opportunity, cacheuse_hidden_to_kv_opportunity, cacheuse_thrash_risk_high, cacheuse_thrash_risk_medium, cacheuse_thrash_risk_near_capacity.

Scheduler/Method Taxonomy

The dataset logs the exact comparative performance of 13 policies:

1. FIFO (fifo): First-In-First-Out baseline.
2. EDF (edf): Earliest-Deadline-First deadline-aware scheduler.
3. WSP (weighted_shortest_processing): Weighted Shortest Processing Time First.
4. LLF (least_laxity_first): Least Laxity First.
5. ESTF (estimated_service_time_first): Shortest Processing Time First variant.
6. Scorpio (scorpio_style_slo_guard): Scorpio-style SLO-guaranteed queue manager (SOSP 2023).
7. vLLM (vllm_style_token_budget): vLLM-style memory-budget scheduler with reemption-on-exhaustion (SOSP 2023).
8. Sarathi (sarathi_style): Sarathi-style chunked-prefill scheduler (OSDI 2024 / CoRR 2023).
9. Orca (orca_style): Orca-style continuous batching iteration-level scheduler (OSDI 2022).
10. Shortest Output First (shortest_output_first): Prioritizes short generation outputs.
11. SLO Slack Score (slo_slack_score): Prioritizes requests based on remaining SLO slack.
12. Admission Control (admission_control): Standard request filtering baseline.
13. APT-Serve (apt_serve_faithful): Preemption-aware, transition-optimal scheduler developed by Soroush Vahidi, evaluated under 5 distinct switch-penalty/transition-cost multipliers (0x, 0.5x, 1x, 2x, 4x).

Data Dictionary

Column	Type	Meaning
stage	string	Evaluation phase: screening or confirmation.
regime_id	string	Traffic envelope scenario ID.
seed	int64	Random traffic distribution seed.
policy_label	string	Evaluated scheduler policy.
policy_kind	string	Categorization (baseline or apt_serve).
transition_cost	string	Transition cost penalty (na for baselines; 0x_idealized, 0.5x, 1x, 2x, 4x for APT-Serve).
kv_pressure	string	KV cache pressure rating.
slo_pattern	string	SLO latency target pattern.
length_pattern	string	Prompts/decodes length distribution.
arrival_pattern	string	Traffic arrival rate distribution.
cache_use_structure	string	Context reusability structure.
num_completed	int64	Number of completed requests.
num_dropped	int64	Number of dropped requests.
num_total	int64	Total requests in workload.
completion_fraction	float64	Completed / total fraction.
arrival_normalized_weighted_goodput	float64	Goodput normalized by arrival.
weighted_goodput_completed_only	float64	Completed-only weighted goodput.
slo_violation_rate	float64	Fraction of completed requests violating the latency SLO.
mean_latency	float64	Mean request latency in seconds.
p95_latency	float64	95th percentile request latency.
mean_ttft	float64	Mean Time-To-First-Token in seconds.
p95_ttft	float64	95th percentile TTFT.
request_throughput	float64	Requests completed per second.
token_throughput	float64	Tokens generated per second.
kv_to_hidden_transitions	Int64	APT-Serve enriched metric: active-to-inactive cache transitions.
hidden_to_kv_transitions	Int64	APT-Serve enriched metric: inactive-to-active cache transitions.
evictions	Int64	APT-Serve enriched metric: block-level cache evictions.
recomputations	Int64	APT-Serve enriched metric: preemption-induced recomputations.
switch_latency_paid	float64	APT-Serve enriched metric: total transition overhead latency.
restore_latency_paid	float64	APT-Serve enriched metric: total cache restoration latency.

Simulator and Environment Assumptions

Metrics are logged from Soroush's high-fidelity event-driven LLM serving simulator.

• GPU models are parameterized with homogeneous memory capacities.
• Preemption is modeled via recomputation or swap-to-host.
• Transition-overhead factors (switch_latency_paid, restore_latency_paid) reflect physical hardware swap penalties under strict bus-width assumptions.

Scientific Use

This dataset allows researchers to perform deep comparative analyses of LLM-serving schedulers under severe traffic pressure, study Pareto throughput/goodput curves, model optimal preemption thresholds, and validate state-of-the-art scheduling policies without re-scaffolding complex event-driven simulations.

Non-Intended Uses and Limitations

The results are simulated performance outcomes under specific mathematical queueing/hardware models and should not be treated as empirical physical GPU serving measurements.

Difference from SoroushVahidi/module-intervention-credit

• SoroushVahidi/module-intervention-credit is specifically designed for AutoML counterfactual credit assignments in modular pipeline networks.
• This dataset is a monolithic baseline scheduler evaluation benchmark comparing 13 full-system scheduling algorithms across 41 extreme traffic envelopes.

Associated Paper and Citation Policy

No directly verified preprint paper is associated with this dataset release yet.

Citation Guidance:

• Cite this dataset when utilizing these simulation rows or benchmarking comparisons.
• Cite upstream scheduler papers (such as vLLM [SOSP '23], Orca [OSDI '22], Sarathi-Serve [OSDI '24], or Scorpio [SOSP '23]) when discussing or relying on the baseline models implemented here.

@dataset{vahidi2026_llm_serving_scheduler_baselines,
  author = {Vahidi, Soroush},
  title = {LLM-Serving Scheduler Baselines: Event-Driven Simulation Performance Outcomes for 13 Scheduling Policies},
  year = {2026},
  version = {v1},
  publisher = {Hugging Face},
  url = {https://huggingface.co/datasets/SoroushVahidi/llm-serving-scheduler-baselines},
  license = {CC-BY-4.0}
}