Table of contents
At a glance
How does IT infrastructure qualification differ from software validation?
Infrastructure qualification assesses whether the required technical foundation is provided and operated under control. Application validation demonstrates fitness of the specific application for its intended use. A qualified server does not prove that batch logic is correct. A successful application test does not replace all controls for backup, access and recovery. The two sets of evidence should connect.
Scope and Applicability
Annex 11 generally distinguishes between application and IT infrastructure. Consequently, a single, identical documentation package is not required for every architecture. Cloud, virtual environments, and on-premise operations each require appropriate system boundaries and responsibilities.
The limit lies in the function, not merely in the supply agreement
A technical baseline facilitates alignment. It describes the intended state and relevant configurations. Not every detail needs to be verified at the same level of depth; however, critical dependencies must be identifiable. Modifications to shared services may trigger an evaluation of multiple applications.
For outsourced services, supplier certifications or service reports should be evaluated. A certificate or service report may support portions of infrastructure validation but does not automatically demonstrate the suitability of your specific configuration. Document which evidence is accepted and which controls are performed in-house.
Infrastructure and domain-specific applications are interrelated
Not every technical device requires the same level of documentation. Describe the infrastructure capabilities required by the specialist application and identify potential outages or malfunctions that could impact its data and processes.
Capture servers, databases, network, storage, identity services, time sources, and other actually required components. For cloud services, part of this is managed by the provider. The responsibility matrix must demonstrate which evidence is available and which tasks remain with the customer. A cloud agreement does not replace this assignment.
Start with the required function: appropriate access, data retention, recovery, or temporal traceability. A complete inventory of all devices forms a foundation, but is not yet proof of suitability. The qualification strategy integrates technical performance with intended use.
Develop a risk-based infrastructure matrix
Name the component, required attribute, responsible party, and testing method. Not every function needs to be documented separately in an IQ, OQ, or PQ document. Existing appropriate operating and testing records may be utilized if the status and statement are compatible. The format shall follow the organization's QMS and the required evidence.
Document common usage. A technical change may impact multiple specialized applications. The impact assessment should identify these dependencies. A server that appears non-critical for one application may play a central role in another process.
| Component | Required statement | Evidence to produce |
|---|---|---|
| Storage | Required data remain available and properly accessible | Appropriate operation and restoration validation |
| Identity service | The right individuals receive appropriate access. | Representative role and failure scenarios |
| Time source | Events have the required time reference | Testing against defined requirements |
| Network | Controlled data pathways are manageable. | Connection, interruption and restart |
| database | The application uses the intended system. | Configuration and appropriate functionality tests |
| Cloud-operated | Supplier and customer responsibilities are covered | Evaluated evidence and internal controls |
Example: Server restart causes the application to fail to locate its data
A server is restored from a backup and is accessible. However, the dedicated application references a file storage location with attachments that has not been restored. The technical startup test is passed, but full functional validation fails. The required infrastructure exceeds a single virtual machine.
Capture database, file storage, identity, and licensing dependencies. Test an appropriate recovery procedure using known reference scenarios. Subsequently, assess application, data relationships, and required access. The scope is determined by architecture and risk; a universal complete system failure scenario for each system is not assumed.
Upon identification of an issue, the restoration procedure description is corrected and re-evaluated. Identify the required sequence and responsible parties. An additional backup update alone does not confirm complete system recovery. The evidence should demonstrate the infrastructure's capability to support the specific application requirements.
| Test case | Action | Expected finding |
|---|---|---|
| Technical Start | Start the server and services | Available basis |
| Data access | Open known process | Correct data set available |
| Attachment | Search external file | Required relationship and content available |
| Identity | Use the specified professional role | Access functions appropriately |
| Interruption | Data pathway monitored and interrupted as required | Detection and retry handling |
| Recovery | Establish full dependence | Use in the process restored |
Supplier evidence and internal inspections combine
A supplier may provide infrastructure tests, security reports, or operational information. Assess which specific capability and which environment are covered. A general certificate does not constitute evidence of any customer-specific restoration or network connectivity. The evaluation should utilize appropriate evidence and identify remaining tasks.
When responsibility is shared, handover points must be clearly defined. The supplier may restore the storage, while the customer reviews the application and technical data. A technical SLA does not automatically address all quality requirements. Identify which decision is based on which feedback.
Avoid duplicate assessments without additional justification. If an appropriate validation evidence is available, it may be utilized. At the same time, the absence of evidence must not be overlooked by referencing "standard infrastructure." The decision shall be based on the required performance and actual risk.
Evaluate changes, patches, and capacity appropriately
An infrastructure change may affect access, performance, or data processing. Review impacted applications and operating conditions. A security patch may impact a limited function or require a broader restart. The technical designation alone does not determine the scope of assessment.
Consider capacity and load, to the extent they are relevant to usage. A test with a small number of data sets and a single user does not validate every expected operating scenario. Define the required performance and an appropriate testing approach. The article does not specify a general response time or universal load value.
Get the relation between tested and productive status. Relevant differences in network, identity or memory are to be evaluated. A successfully qualified platform may lose its previous statement due to later uncontrolled change. Change control and ongoing evaluation remain connected.
Operate and handover based on specific responsibilities.
Name responsible parties for monitoring, security, recovery, and failure. Technical and professional assessments should complement each other. A green infrastructure monitor does not automatically confirm accurate business data. Conversely, a professional error may require a technical investigation.
Provide architecture, dependencies, validation evidence, and restart procedure. A practical exercise based on a well-known process demonstrates whether the team identifies and understands the required steps. The handover should also include coverage of representation and external suppliers.
Upon system decommissioning, it must be assessed whether any infrastructure is still required to support historical access. An older licensing or identity service may continue to support archival functions. This dependency must not be terminated without notice or without proper evaluation when transitioning to another application. The data lifecycle links qualification, operation, and a structured exit strategy.
Draft: Assess required infrastructure performance up to point of use
Select a specific performance metric such as recovery after failure. Describe the complete set of required components. In the example, the database, attachments, identity, and license are interdependent. A technical server restart only validates a portion of this chain. Professional usage should be verified against established reference procedures.
Assign provider and customer responsibilities unambiguously. The cloud provider may offer storage or recovery services, while the customer is responsible for validating applications and data context. The matrix should clarify such handover points. A general SLA or certificate is not a substitute for the actual required overall performance.
An investigation is conducted to determine the root cause and other affected applications. Shared infrastructure may encompass a broader scope than the initially assessed case. Additional evaluation should take into account this dependency and utilize appropriate validation evidence, without necessarily performing a full, independent revalidation of each individual function.
| Performance area | Example entry | What the test establishes |
|---|---|---|
| Application | Open historical inspection task. | Required professional function. |
| database | Designated restored status. | Correct data available. |
| File storage | Associated attachments. | No lost external connection. |
| Identity | Designated process role. | Access and boundaries function. |
| Dependency | Required license or key service. | Maintain technical usability. |
| Transfer | Responsible Roles and Retry Sequence. | Operation can practically execute the pathway. |
Next steps
- Assign required infrastructure services and applications.
- Define technical specifications and responsibilities.
- Verify critical supply and operational functions.
- Evidence for common services controlled for reuse.
Project experience and collaboration
Trace technical recovery through to the business process
With operations and application owners, we test whether a restored environment supports the required transaction. Server startup is an intermediate step. Access, data connections and usable business records all need to work together. Existing evidence may cover shared services already under control. Changed dependencies and previously untested recovery routes need focused additions. This defines a qualification scope grounded in application operation.
Frequently asked questions about IT infrastructure qualification
Must each network device be individually extensively qualified?
What does the cloud provider assume responsibility for, and what remains the customer's responsibility?
The supplier may provide essential technical capabilities and supporting documentation. Your organization must assess the suitability of such capabilities for its own regulated use and assume any remaining responsibilities. Configuration, access controls, interfaces, and professional usage may require independent validation evidence. The specific allocation of responsibilities will be clearly defined in the operating model and the agreed arrangements.
Does each infrastructure component require a separate IQ/OQ/PQ package?
The required evidence depends on the task, risks and existing QMS. Suitable existing operating, installation and function proofs can be used if their statement fits. The decisive factor is the comprehensible suitability of the required infrastructure performance, not the mere number of separately-titled documents.
How do we differentiate infrastructure qualification and application testing?
The infrastructure assessment confirms the required technical capabilities and conditions. The application test verifies the functional performance in the intended usage scenario. Both can complement each other. A reachable server does not prove correct computation; a correct individual value does not automatically demonstrate complete recoverability or an appropriate shared platform.
Why must external file storage be present in the recovery test?
Since business records may reference attachments or other content outside the database, a restored primary server may start up while these information elements are missing. Review known complete processes. The required scope aligns with the actual architecture and technical justification.
Sources and review date
The sources provide the technical foundation. Tables, checklists, and examples are derived methodological tools, not verbatim regulatory requirements.
- European Commission: EU GMP Annex 11, Computerised Systems (2011)Principles as well as §§ 1, 3, 4.7, 7, 10 and 16: Application, Infrastructure, Responsibilities, Operation and Continuity. Current version as of EudraLex on 14.10.2026; no universal project or deadline specification.
- European Commission: EU GMP Annex 15, Qualification and Validation (2015)Qualification and validation, supplemented by the specific framework for computerized systems.
Sources checked: .
From an open point to a traceable approach.
We support you in assessing the current situation, clarifying responsibilities, and implementing appropriate evidence within operations. The type of support that is most effective depends on your system, your team, and the specific objective.
- Computer system validation →
Bring your GxP systems into validated operation, from laboratory software, MES and eQMS to ERP. We define the scope, plan risk-based tests using GAMP 5 and hand the evidence to QA for release.

