Search arXiv⌕ Search
Home/Archive/arXiv

arXiv · 0705.2065

Mean Field Models of Message Throughput in Dynamic Peer-to-Peer Systems

Aaron Harwood·Olga Ohrimenko
View original source ↗Read source document ↗Export metadata ↓

Abstract

The churn rate of a peer-to-peer system places direct limitations on the rate at which messages can be effectively communicated to a group of peers. These limitations are independent of the topology and message transmission latency. In this paper we consider a peer-to-peer network, based on the Engset model, where peers arrive and depart independently at random. We show how the arrival and departure rates directly limit the capacity for message streams to be broadcast to all other peers, by deriving mean field models that accurately describe the system behavior. Our models cover the unit and more general k buffer cases, i.e. where a peer can buffer at most k messages at any one time, and we give results for both single and multi-source message streams. We define coverage rate as peer-messages per unit time, i.e. the rate at which a number of peers receive messages, and show that the coverage rate is limited by the churn rate and buffer size. Our theory introduces an Instantaneous Message Exchange (IME) model and provides a template for further analysis of more complicated systems. Using the IME model, and assuming random processes, we have obtained very accurate equations of the system dynamics in a variety of interesting cases, that allow us to tune a peer-to-peer system. It remains to be seen if we can maintain this accuracy for general processes and when applying a non-instantaneous model.

Explore related subjects

cs.DCcs.PF

Keep this discovery

Explore connections, maps & timelines

BibTeXRIS

Aaron Harwood, Olga Ohrimenko. 2007-05-15. Mean Field Models of Message Throughput in Dynamic Peer-to-Peer Systems. https://arxiv.org/abs/0705.2065

Cite the original work for its findings. Save a collection to share your selection of sources.

KEEP EXPLORING

Related papers

arXiv2026

Scepsy: Serving Agentic Workflows Using Aggregate LLM Pipelines

Agentic workflows carry out complex tasks by orchestrating multiple large language models (LLMs) and tools. Serving them at a target throughput with low latency is hard because they are written in arbitrary agentic frameworks and their execution times are unpredictable: execution branches, fans out, or recurs in data-dependent ways. Since their LLMs often outnumber the available GPUs, they also oversubscribe GPUs. We describe Scepsy, a serving system that schedules arbitrary multi-LLM agentic workflows onto a GPU cluster. Scepsy exploits the insight that, while the end-to-end latency of an agentic workflow is unpredictable, each LLM's fraction of execution time is comparatively stable across requests. Scepsy profiles each LLM under different parallelism degrees and combines the profiles with these fractions into an Aggregate LLM Pipeline, a lightweight throughput and latency predictor for allocations. To minimize latency at a target throughput, Scepsy uses the Aggregate LLM Pipeline to search over fractional GPU shares, tensor parallelism degrees, and replica counts. A hierarchical heuristic then places the chosen allocation onto the cluster, minimizing fragmentation and respecting network topology. On realistic agentic workflows, Scepsy achieves up to 2.5x higher throughput before saturation and 1.0-3.3x lower latency than systems that optimize LLMs independently or rely on user-specified allocations.

cs.DC↗
arXiv2026

Bandwidth-Aware and Cost-Efficient Pipeline Parallel Scheduling in Geo-Distributed LLM Training

The rapid evolution of large language models (LLMs) has made geographically distributed training necessary due to GPU scarcity within a single cloud region. In such cross-region settings, Pipeline Parallelism (PP) is communication-efficient, yet scheduling PP remains challenging under heterogeneous inter-region bandwidth and regional electricity prices. Existing schedulers are either delay-first, incurring high electricity cost, or cost-first, relying on rigid resource allocation that prolongs Job Completion Time (JCT). They are also ineffective at optimizing execution order in multi-tenant environments, where long-running and bandwidth-intensive jobs can cause head-of-line (HoL) blocking and degrade overall performance. To this end, we propose BACE-Pipe, a bandwidth-aware and cost-efficient pipeline scheduling framework for LLM training across geo-distributed clusters. BACE-Pipe first introduces a dynamic job prioritization mechanism that optimizes execution order by jointly considering job characteristics (e.g., computation time) and real-time network utilization. It then employs a bandwidth-aware pathfinder to identify feasible cross-region pipeline paths that satisfy communication constraints, thereby preventing communication from stalling the pipeline. Among all feasible paths, a cost-minimizing allocator determines the optimal GPU placement strategy by preferentially assigning resources to regions with lower electricity prices. Consequently, BACE-Pipe mitigates HoL blocking, improves resource utilization, and simultaneously reduces both JCT and total electricity cost. Extensive simulations show that BACE-Pipe reduces average JCT by 27.9%--64.7% and total electricity cost by 12.6%--30.6% compared with state-of-the-art baselines.

cs.DC↗
arXiv2026

WeEnv: The Environment for Agentic Reinforcement Learning at WeChat

Agentic reinforcement learning (RL) differs from conventional RL in that every task executes inside a complex environment, e.g., a virtual machine or a container. We find that agentic RL pays a heavy environment tax: a large share of the iteration time goes to the environment rather than to learning. The root cause is the lack of a full-lifecycle solution to environment management. We present WeEnv, which manages environments across packaging, initialization, and provisioning. WeEnv packages components as independently published layer groups and composes them at initialization, so that updating a component republishes one small group rather than every artifact containing it. To speed up environment initialization, WeEnv launches environments instantly and fetches contents on demand. During task execution, WeEnv provisions CPU and memory elastically, adjusting each environment's quota from its observed usage to fit the varying demands. WeEnv reduces the initialization by 5.6-14.2x over E2B, Docker, and AgentENV, cutting its share of the iteration time from up to 53.4% to 9.1%. WeEnv is deployed for agentic RL at WeChat.

cs.DC↗