Resource contention
Resource contention is competition among concurrent operations, processes, or workloads for access to a limited shared resource.
Latency
Resource contention is competition among concurrent operations or workloads for a limited shared resource, causing some to wait.
- How it is used
- Explains latency caused by competition for shared resources across virtual machines, locks, and concurrent requests.
- How deep it goes
- Defines waiting mechanisms, relates concurrency to latency with Little’s law, and compares lock timings as thread counts rise.
- Examples
- Uses a shared-counter mutex benchmark and virtual machines competing for CPU time.
- What you can do after
- Diagnose contention by comparing lock timings across thread counts and measuring application latency under different partitioning strategies.
C++ Concurrency in Action, Second Edition
Thread contention is competition among threads for access to shared data, a resource, or a synchronization primitive.
- How it is used
- Uses contention to reason about multithreaded performance and guide data-structure and work-queue design.
- How deep it goes
- Explains waiting, cache transfers, and cache ping-pong, then discusses layout and queue designs that reduce contention.
- Examples
- Uses a shared atomic counter, a mutex loop, per-thread work queues, and a shared stack of pending work.
- What you can do after
- Diagnose performance risks by identifying shared counters, mutex loops, queues, and stacks that can become contended.
Read how C++ Concurrency in Action, Second Edition explains resource contention →
Kubernetes for Developers
Resource contention is competition among workloads for finite computing resources on shared infrastructure.
- How it is used
- Explains how Kubernetes preserves node operation when Pods compete for finite CPU or memory.
- How deep it goes
- Connects requests, limits, QoS, priority, eviction, and rescheduling, but stops short of detailed capacity-sizing procedures.
- Examples
- Uses timeserver Deployments, CPU and memory settings, QoS classes, eviction-order figures, and priority-class YAML.
- What you can do after
- Diagnose frequent evictions and configure requests, limits, QoS, priority, and cluster capacity accordingly.
Read how Kubernetes for Developers explains resource contention →
Grokking GPU Programming
Contention occurs when multiple threads compete to access or update the same resource, potentially limiting parallel performance.
- How it is used
- Explains why conflicting updates can limit performance in parallel histograms.
- How deep it goes
- Traces atomic serialization through hot-bin slowdowns and describes block-local privatization, without detailed hardware analysis.
- Examples
- Uses a mostly black image, identical values, and evenly distributed histogram bins to illustrate differing contention.
- What you can do after
- Apply CUDA atomics or block-local histogram privatization to handle competing updates and reduce global contention.
Read how Grokking GPU Programming explains resource contention →
Also covered in
What to know first
- atomicityAtomicity is the property that an operation or group of operations takes effect as an indivisible whole, without observable partial effects.
- concurrencyConcurrency is the structuring or execution of multiple computational activities so their progress overlaps in time, whether or not they run simultaneously.
- GPU memoryGPU memory is memory available to a graphics processing unit for storing data and state used in graphics or computation.
- performancePerformance is a measure of how effectively a person, process, model, or system achieves its objectives, including qualities such as output quality, speed, and resource efficiency.
Related concepts
- append-only systemaddressed by · A system that records new data by adding entries while preserving existing entries without modifying or removing them.
- mutexuses · A mutex is a synchronization primitive that provides controlled exclusive access to shared data or a protected operation, admitting only one thread at a time.
- volatilityproduced by · Volatility is the degree or frequency with which a quantity, condition, or system varies or changes over time.
- meanproduces · The mean is a measure of central location equal to the arithmetic average of numerical observations, or the probability-weighted average of a random variable’s possible values.
- resource limitaddressed by · A resource limit is an enforced upper bound on the amount or rate of a resource that an entity may consume within a defined scope or period.
- shared mutable stateproduced by · Shared mutable state is data or computational state that multiple components or execution contexts can access and modify.
Questions about resource contention
What is resource contention?
Resource contention is competition among concurrent operations, processes, or workloads for access to a limited shared resource.
Which Manning books explain resource contention?
Latency, C++ Concurrency in Action, Second Edition, Kubernetes for Developers and Grokking GPU Programming.
What should I know before learning resource contention?
Manning books that teach resource contention usually introduce atomicity, concurrency, GPU memory and performance first.
These guides are generated from the text of Manning books by Manning's knowledge graph and link to the sections they draw on. The full books are available on liveBook and at manning.com.



