IT systems · Practical guide

SAP IDoc validation: retries and the resulting document

The sending system reports a connection error after SAP has already posted the goods movement. Recreating the IDoc may cause a second posting. Define which retry is safe and show that the intended process state is reached exactly once.

Table of contents

At a glance

What should you test when validating a SAP IDoc interface for GxP?

Test the entire path from the source business event to the target document. Define identity, mapping, sequence, error handling and permitted retries. Technical IDoc or RFC status alone does not establish correct processing. Include lost acknowledgements, new message identifiers, partial processing and manual rework. Reconcile the final data and business state.

The technical notification is more than an IDoc

An IDoc structures a data exchange. For validation, identify the business action that triggers transmission. Does it create an inspection order, block a batch or post a goods movement? Describe the source event, the expected outcome in the target and the required response. A formally valid message can affect the wrong material, plant or batch. Technical receipt leaves those assignments unverified.

Capture message type, base type, extensions, partner agreements, and relevant processing components aligned with the actual architecture. Middleware, source system, and customer-specific processing should be considered where they impact the outcome. Not every SAP edition employs the same integration pathway. This work guide is intended for environments where IDoc processing is actually in use. It does not claim universal IDoc availability or configuration for every SAP product.

Sources: [1]

Transport identity and business identity are distinct.

SAP defines for tRFC a one-time execution within its transaction mechanism, and for qRFC an explicitly defined sequence within the corresponding queue. The transaction ID, TID, serves a specific technical function. This characteristic must not be transferred without verification to any end-to-end chain. A preceding system may resend the same business action under a new identity. In this case, the technical message is new, even though the intended booking already exists.

Therefore, define an appropriate professional identity, for example, a combination of source system, document number, position, and event version. Which repetitions should be considered part of the same process? When does a later event constitute a legitimate correction? A duplicate criterion based solely on material and quantity would often be too coarse: two identical deliveries could both be valid. A criterion based on a newly generated message number each time could, in contrast, allow the same delivery to be recorded multiple times. The rule is derived from the process.

Sources: [2]

The target document bears the actual proof

Reconcile the target document and its identity, quantity, unit, batch, status, date and references alongside the IDoc count and transmission status. Check whether the document was posted or only an intermediate step exists. Technical processing can finish while a subsequent business action is still outstanding. State what process state has actually been reached.

Feedback also requires a clear link to its source event. A message indicating "success" without a unambiguous connection to the originating event provides little value in scenarios involving multiple concurrent processes. Define clearly which information the sender provides to support a decision on whether to retry. Describe whether professional judgment, transport error, and an ongoing processing status can be distinguished. A timeout initially signifies only that the expected response was not received in a timely manner; it neither proves the failure nor the success of the target action.

Example: Approval lost, booking exists

In its own testing scenario, a warehouse initiates a 20 kg entry for Batch A. SAP generates the corresponding target booking. Subsequently, confirmation in the controlled test track is interrupted. The sender observes no success. The test now evaluates the approved repeat rule: a technical re-attempt of the same action must not inadvertently generate an additional entry exceeding 20 kg. The expected inventory level and booking count are determined independently of the interface log.

In addition, generate the same business content using a new message identity. This enables testing of a different error class than the repetition of the same transport transaction. The desired response may be an identified duplicate, a targeted follow-up action, or another specified handling procedure. Do not assert that SAP automatically prevents all business-level duplicates. The following matrix outlines proposed validation scenarios. Execution and intervention take place in a separately authorized test environment, not through manipulation of actual production postings.

Own work aid – Example: Confirmation lost, booking exists
TriggerPre-defined expectationsTechnical Alignment
Release confirmation after successful booking is missing Controlled clarification or reprocessing without a second effect An expected access, correct documentation reference
Same business action, new message ID Duplicate rule is logically traceable No undetected extra posting
Correction after a previously processed event Correction remains distinguishable from the original Correct final balance and complete history
The block and subsequent release arrive out of order Specific sequence rule remains effective No unauthorized temporary usage state
One of several positions is invalid Defined overall or partial reprocessing status Reconcile every line item
Manual follow-up processing Justified, traceable action Relationship between error, change, and intended effect

Check sequence according to technical dependency

Not every message requires a global ordering. However, event dependencies may necessitate it: create batch, verify, lock, and later release. Describe the dependency and evaluate the employed queue or ordering mechanism. The qRFC property applies within its specific mechanism. Different queues, middleware paths, or concurrently triggered business events are not automatically consolidated into a single, functionally ordered sequence.

Test delay, congestion, and restart scenarios. Can an older event occur after a newer state and invalidly overwrite it? Is version or state verification required? When multiple systems are involved, validation evidence must demonstrate the effective state at each relevant location. A batch appearing locked in the ERP system but still being issued from inventory does not constitute a sufficiently substantiated end-to-end lock process. The test must also consider the period until full effectiveness is achieved.

Sources: [2]

Manage partial processing and targeted modifications

A message package may contain multiple positions. Determine whether a failure resets the entire processing or whether partial effects remain acceptable. Evaluate this decision specifically. Following a retry, previously processed positions must not execute twice, and previously erroneous positions must not be lost. A final status for the entire package must have a clear reference to all required sub-actions. Otherwise, a summary notification may obscure open positions.

When changing mapping, master data or processing logic, you look at existing open messages. Are they processed according to the new or the old rule? Can a new field assignment change the meaning of historically waiting messages? Describe how versions and post-processing are assigned. A new successful test with freshly generated messages does not automatically cover the treatment of an old residue. The controlled transition belongs to the change concept if such residues are present.

Post-processing and operation inclusion

Manual post-processing may be technically necessary. Define responsibilities, permitted corrections, and release boundaries. Which changes remain traceable? Is a modified message reprocessed, or is a replacement procedure established? Verify that the original content and the relationship to the correction are preserved to the required extent. Administrative access enabling arbitrary repetitions without clear technical justification requires appropriate evaluation and control.

A useful monitoring system distinguishes open, rejected, processed and ineligible processes. Those responsible should be able to identify which technical decisions are affected. A green connection message does not replace this information. Determine escalation and reaction after process effect. For a time-critical blocking process, the permissible response time may be different from for a later statistical evaluation. The deadline is professionally justified and not taken over from a standard technical number.

Sources: [1] [3]

Create a traceable final record

The final image integrates source event, transport identity, processing, target assignment, and feedback. Document the error pathways that have been evaluated and those that remain unaddressed. This includes behavior under conditions of lost confirmation, professional duplicate, reversed sequence, partial processing, and retry. Record allowable differences and the rationale behind them. A simple comparison of "100 sent, 100 received" does not address these scenarios when one message is processed twice and another is entirely omitted.

Provide the operations team with clear search and decision pathways. The team must be able to resolve ambiguous states without repeatedly initiating redundant transmissions. A work instruction should specify which references to compare and when professional support is required. Modifications to interface specifications, identity rules, or operating procedures trigger an appropriate impact assessment. Thus, the original validation evidence remains applicable after updates, provided its underlying conditions continue to be met.

Project experience and collaboration

Link the IDoc processing attempt to the resulting business document

With process owners and the interface team, we trace a known source transaction to the correct target document. Controlled cases cover an interrupted acknowledgement, a repeated message and partial processing. Where the architecture supports unambiguous matching and restart, we can test those controls directly. Other cases may need further reconciliation or a more tightly controlled manual investigation. We record exactly what the IDoc status confirms and which business checks remain necessary.

Frequently asked questions about Validating SAP IDoc processing

Does a successful IDoc status confirm correct GxP booking?

The status designates a processing step. In classical ALE inbound processing, SAP, for example, assigns value 53 to "application document booked" and 51 to "application document not booked". Even a booked document must still be correctly assigned to the appropriate business object with the correct content. Review status, application document, and actual destination usage jointly. Root cause analysis and corrective actions must align with the specific procedure; repetition may impact previously performed partial processing.

Sources: [4] [1]

Does tRFC prevent any duplicate business transaction?

The SAP property concerns the one-time execution within the intended transaction mechanism with its identity. A re-created business action with a new technical identity is another test subject. Therefore, evaluate professional duplicates and the entire transmission chain additionally. From a transport property no universal duplicate freedom follows.

Sources: [2]

Does a timeout indicate that we should immediately resend?

No. A timeout does not confirm the actual target state initially. The process may have already been processed. Apply the defined clarification and retry rules and compare the relevant references. A blind-generated replacement process may trigger a second impact on the process.

Does a qRFC-Queue suffice for each sequence guarantee?

Verify the actual queue assignment and technical dependencies. The described sequence applies within the employed mechanism. Other queues or upstream paths may require additional evaluation. Specifically test delayed older events and restart behavior following processing backlog.

Sources: [2]

Do we always need to develop a complete in-house duplicate testing procedure?

Not at all. First, examine existing mechanisms and their actual coverage. The process must master the relevant errors; standard functions, configuration or additional measures can be suitable for this. Crucial is a reliable proof of your concrete identity and repetition rules, not a certain level of development.

Which documents belong in the interface test?

The following are required: traceable requirements, inputs, message and action references, relevant status information, and the actual observed outcome. Risk-based evaluation of errors and reoccurrence incidents should be included. The evidence should enable independent assessment, without relying solely on a general success report from the transfer tool.

Sources: [1]

Sources and review date

Status: October 15, 2026. Regulatory classification and manufacturer responsibilities are separately documented. Examples, test matrices, and templates are standalone reference tools; they are customized for actual usage purposes.

  1. European Commission: EU GMP Annex 11, Computerised Systems (2011)Current version as per EudraLex inventory as of 15.10.2026. Application, risks, operation, and data lifecycle; specific testing proposals are derived independently.
  2. SAP Library: tRFC and qRFC SupportDocumentation on the Connector Principle: TID, one-time implementation, qRFC sequence. No evidence of an arbitrary middleware or business process chain.
  3. PIC/S: PI 041-1, Good Practices for Data Management and Integrity (2021)Inspection Guidance on Data Integrity; in particular electronic data, transmission, access, and review.
  4. SAP Library: Testing Inbound Processing – classical IDoc/ALE ProcessingOld SAP library documentation: incoming status examples 53/51 and verification of actually written application data. Validate own message type and processing procedure.

Sources checked: .

Daniel Herrmann, Daniel Herrmann Consulting

Daniel Herrmann Consulting

How we can support you

Daniel Herrmann Consulting combines professional assessment with hands-on work on the specific system. During the first meeting, we clarify your intended use, existing documentation, and the next most appropriate action step.

  • SAP S/4HANA validation →

    Support an SAP go-live with evidence your team can defend at audit while keeping validation on schedule. We take a risk-based approach from the template to the interface for greenfield or brownfield projects and existing ECC systems.

Send a project enquiry Customer testimonials →