Vendor lock-in
Vendor lock-in is dependence on a provider, product, or technology that makes switching to an alternative difficult, costly, or impractical.
Kafka for Architects
Vendor lock-in is a condition in which proprietary dependencies, integrations, or switching costs make migration away from a provider difficult or impractical.
- How it is used
- Guides enterprise decisions between managed cloud Kafka and on-premises deployment.
- How deep it goes
- Explains proprietary dependencies, configuration limits, egress costs, and portability mitigations without detailing migration procedures.
- Examples
- Uses Confluent Cloud, Amazon MSK, Aiven, proprietary integrations, egress fees, and dual Kafka clusters as examples.
- What you can do after
- Evaluate provider portability, document exit costs, and preserve migration options with standard serializers and APIs.
Vibe Engineering
Vendor lock-in is dependency on a particular supplier that makes switching to an alternative costly or difficult.
- How it is used
- Reduces migration risk by isolating AI-provider choices behind configuration and a replaceable application boundary.
- How deep it goes
- Demonstrates provider substitution operationally, but stops short of a broad economic analysis of switching costs.
- Examples
- Uses Gemini and OpenAI switching, Cloud Run updates, provider logs, API keys, and generated SQL validation.
- What you can do after
- Operate/configure a deployed service to switch AI providers while preserving application logic and checking runtime initialization.
Designing AI Systems
Vendor lock-in is dependence on a provider's technology or interfaces that makes switching to alternatives difficult or costly.
- How it is used
- Reduces provider dependence by placing model integrations behind a shared Model Service interface.
- How deep it goes
- Explains operational causes, organizational costs, normalization needs, and adapter-based mitigation without providing implementation code.
- Examples
- Uses OpenAI, Anthropic, and Google Gemini formats, plus an outage fallback and token-usage normalization scenario.
- What you can do after
- Evaluate provider choices and apply a Model Service with adapters to switch providers without rewriting application logic.
First-Party Data Activation
Dependence on a technology or vendor that makes replacing, migrating from, or interoperating with it difficult or costly.
- How it is used
- Evaluates first-party data solutions by their portability, offboarding difficulty, and effect on marketing-stack agility.
- How deep it goes
- Explains business-logic portability and compares solution categories, but stops short of implementation procedures or contractual analysis.
- Examples
- Uses audience criteria, personalization fields, experiment configurations, and audience membership lists as portability examples.
- What you can do after
- Evaluate vendors by asking how business logic and data can be offboarded and whether downstream systems can change independently.
Read how First-Party Data Activation explains vendor lock-in →
The Cloud at Your Service: The when, how, and why of enterprise cloud computing
Vendor lock-in is dependence on a provider or its technologies that makes switching to an alternative difficult or costly.
- How it is used
- Frames provider dependence as a trade-off when choosing cloud platforms and deployment approaches.
- How deep it goes
- Compares provider-specific constraints and portability, but does not detail switching procedures or mitigation plans.
- Examples
- Contrasts App Engine, Azure, EC2, Ruby on Rails, Eucalyptus, and OpenNebula as examples of differing provider dependence.
- What you can do after
- Evaluate cloud choices by weighing provider dependence against scaling, time-to-market, and portability.
Architecting an Apache Iceberg Lakehouse
Tool lock-in is dependence on a particular software tool or vendor that makes changing tools difficult or costly.
- How it is used
- Guides lakehouse architecture and tool selection by exposing migration, interoperability, cost, and operational tradeoffs.
- How deep it goes
- Explains proprietary formats and modular alternatives, then evaluates operating tradeoffs; stops short of migration procedures.
- Examples
- Uses proprietary warehouses, Iceberg lakehouse layers, standard APIs, managed catalogs, and self-managed catalogs as comparison points.
- What you can do after
- Evaluate architecture and tool choices for interoperability, operational burden, cost growth, and dependence on a vendor ecosystem.
Read how Architecting an Apache Iceberg Lakehouse explains vendor lock-in →
Think Like a CTO
Vendor lock-in is a condition in which dependence on a provider makes it difficult or impractical to move software or services to another provider.
- How it is used
- Guides technology decisions toward preserving future options and avoiding dependence that removes practical corrective action.
- How deep it goes
- Defines lock-in, explains its cost trade-offs, and applies a portability test; it stops short of prescribing a migration plan.
- Examples
- Uses obsolete Flash administration software, Java portability, and Salesforce customization to expose dependence and escape options.
- What you can do after
- Evaluate technology choices by testing portability, licensing exposure, downtime consequences, and the cost of preserving alternatives.
Rearchitecting LLMs
Vendor lock-in is a condition in which dependence on a provider makes switching to an alternative prohibitively costly or difficult.
- How it is used
- Frames provider dependence as a production risk when selecting, fine-tuning, and maintaining language models.
- How deep it goes
- Explains financial and technical switching costs through hosted weights, provider lifecycle changes, pricing, and repeated fine-tuning.
- Examples
- Uses fine-tuned closed models, provider upgrades or retirements, pricing discounts, and the leased-house analogy.
- What you can do after
- Operate/configure version pinning and repeat fine-tuning with preserved datasets and evaluations when providers change models.
Also covered in
- Investing for Programmers
- The Creative Programmer
- Cloud Observability in Action
- LLM Customization and Fine-TuningIn early access (MEAP)
- Grokking Software ArchitectureIn early access (MEAP)
- Software Architecture Decision-MakingIn early access (MEAP)
- Bootstrapping Microservices, Second Edition: With Docker, Kubernetes, GitHub Actions, and Terraform
- Learn Docker in a Month of Lunches, Second Edition
- Road to Kubernetes
- Effective Platform Engineering
- ScyllaDB in Action
- Go in Practice, Second Edition
What to know first
- cloud providerA cloud provider is an organization or service operator that supplies remotely accessible computing resources or services over a network, often on demand.
- costCost is the monetary expenditure or other resource burden incurred to acquire, produce, operate, or perform something.
- language modelA language model is a computational model that represents patterns in linguistic sequences to assign probabilities, predict, generate, or evaluate language.
- machine learningMachine learning is a field of computational methods in which systems learn patterns or relationships from data to make predictions, decisions, or other outputs.
Related concepts
- abstractionaddressed by · Abstraction is a simplified representation that preserves relevant properties or behavior while omitting details unnecessary at a given level or for a given purpose.
- model adaptationcontrasts with · Model adaptation is changing a model’s behavior or parameters in response to new data, context, feedback, or a target task or domain.
- open standardaddressed by · An open standard is a publicly available technical specification developed through an open process and implementable by independent parties without exclusive control by a single vendor.
- Apache Icebergaddressed by · Apache Iceberg is an open table format for managing large analytical datasets with metadata, schema evolution, snapshots, and transactional operations across distributed or object storage.
- open-source modeladdressed by · An open-source model is a machine-learning model whose relevant artifacts, such as code or weights, are made available under terms permitting specified forms of access, use, modification, or redistribution.
Questions about vendor lock-in
What is vendor lock-in?
Vendor lock-in is dependence on a provider, product, or technology that makes switching to an alternative difficult, costly, or impractical.
Which Manning books explain vendor lock-in?
Kafka for Architects, Vibe Engineering, Designing AI Systems, First-Party Data Activation, The Cloud at Your Service: The when, how, and why of enterprise cloud computing, Architecting an Apache Iceberg Lakehouse, Think Like a CTO and Rearchitecting LLMs.
What should I know before learning vendor lock-in?
Manning books that teach vendor lock-in usually introduce cloud provider, cost, language model and machine learning first.
What is the difference between vendor lock-in and model adaptation?
Vendor lock-in is dependence on a provider, product, or technology that makes switching to an alternative difficult, costly, or impractical. Model adaptation is changing a model’s behavior or parameters in response to new data, context, feedback, or a target task or domain.
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.







