Monitoring Integration Flows on EIC

Objective

After completing this lesson, you will be able to describe EIC Message Processing Logs, monitoring tools, and explain integration with enterprise monitoring platforms.

Introduction

In Unit 3, Lesson 3.4, you learned that monitoring an EIC deployment spans three layers: integration flow execution (SAP Integration Suite Cockpit), EIC component health (Cockpit + Kubernetes tooling), and Kubernetes infrastructure (customer-managed tools such as Prometheus and Grafana). This lesson focuses entirely on the first layer: the SAP Integration Suite Cockpit as the primary monitoring interface for integration developers.

The key insight: in a hybrid EIC deployment, message processing happens in your private infrastructure, but monitoring remains centralized in the SAP Integration Suite Cockpit in the cloud. This is the operational expression of EIC's control-plane/data-plane separation introduced in Unit 2: the data plane executes locally; the control plane observes and governs from the cloud.

Note

This lesson does not cover Kubernetes infrastructure monitoring (Prometheus, Grafana, kubectl). Those topics were addressed in Unit 3, Lesson 3.4, and belong to the platform operations lay

Monitoring in a Hybrid EIC Deployment

Checking EIC Instance Status in the Cockpit

The first operational check when working with EIC is confirming that the EIC instance is healthy and connected. A disconnected EIC instance means no new deployments can be sent, no configuration updates applied, and no monitoring data received - even though already-deployed integration flows may continue to run locally.

To check EIC instance status:

  1. Navigate to Settings → Runtime → Edge Integration Cell in the SAP Integration Suite Cockpit.
  2. The list shows all registered EIC instances with their current status: Connected, Disconnected, or Error.
  3. A Connected status confirms that Edge Lifecycle Management has an active connection to the Edge Node and the communication channel to SAP BTP is operational.

The Monitor View: Your Primary Operations Interface

The Monitor section of the SAP Integration Suite Cockpit is the single point of access for all integration monitoring - whether flows run on the cloud runtime or on an EIC instance. Navigate to: Monitor → Integrations and APIs

Select Edge Integration Cell Runtime

The Monitor section contains several tiles relevant to EIC operations:

  • Message Processing - view the status of all messages processed by your integration flows
  • Manage Integration Content - see which integration flows and API proxies are deployed
  • Connectivity Tests - verify that adapter connections are reachable from the EIC runtime
  • Security Material - manage credentials and certificates used by your iFlows

The Runtime Filter: The Key Navigation Step for EIC Monitoring

The most important navigation concept for EIC monitoring is the Runtime filter, available in every Monitor tile. By default, the Monitor view shows data from the standard Cloud Integration Runtime. To switch to your EIC instance:

  1. In any Monitor tile, locate the Runtime dropdown (typically in the top-left filter area).
  2. Select your EIC runtime by name - for example, 'FRA-EIC-PROD' for Global Manufacturing AG's Frankfurt instance (the name assigned during provisioning).
  3. All tile data - message counts, deployed artifacts, connectivity status - now reflects your EIC instance only.
Screenshot showing the Runtime drop down used to select an EIC instance for monitoring.

Message Processing Logs (MPL)

Message Processing Logs (MPL): The Audit Trail for Every Message

Every message processed by an integration flow generates a Message Processing Log (MPL). The MPL is the primary artifact for understanding what happened to a specific message: it records the outcome, timing, adapters involved, and - when trace-level logging is active - the step-by-step execution inside the iFlow.

What an MPL Contains

  • Status - the final outcome of message processing
  • Timestamps - when the message was received, when processing completed, and how long it took
  • Adapter details - which sender adapter received the message; which receiver adapter delivered it
  • Processing steps - step-by-step trace of iFlow execution (if trace mode is active)
  • Error details - if the message failed: error type and description of the failure point
  • Correlation IDs - identifiers that link related messages across iFlows

MPL Statuses and Their Meaning

StatusMeaningTypical action
CompletedMessage was processed and delivered successfully.No action needed.
FailedMessage processing ended with an error before successful delivery.Review error details; determine whether to retry or escalate.
RetryThe iFlow encountered a transient error and is automatically retrying.Monitor; alert if retries exhaust without success.
EscalatedThe maximum number of retries was exceeded.Investigate root cause; manual intervention typically needed.
CancelledThe iFlow was stopped while processing this message.Typically occurs during deployment or restart.

EIC - Specific MPL Behavior

MPLs in EIC work identically to the Cloud Integration Runtime from a developer's perspective - the same Cockpit view, the same statuses, the same trace navigation. However, two EIC-specific behaviors are important:

  • Local storage first: MPLs are generated by both the Worker and the Policy Engine. They write log entries asynchronously to the Message Service, from where the monitoring data consumer persists them to EIC's local database (PostgreSQL or SAP HANA). They are not transmitted to SAP Business AI Platform. When you view the Message Monitor in the SAP Integration Suite Cockpit, the monitoring UI sends an API call through the Cloud Connector to query the local EIC API - the data is only displayed to the user on demand, not persisted in the cloud.
  • Offline buffering: If the Cloud Connector loses connectivity to the SAP Integration Suite control plane, message processing continues locally. MPLs are buffered in the local database and transmitted to the Cockpit once connectivity is restored. MPL history is not lost during temporary disconnections, but real-time monitoring in the Cockpit is unavailable during that period.

Only monitoring metadata - status, timestamps, error codes - crosses the network boundary to the cloud. The business message payload remains within your private infrastructure.

Note

Important : When Trace mode is active, message payload data may be transmitted to the SAP Integration Suite Cockpit for step-by-step debugging. This means the data-residency principle - that business payload never leaves your private infrastructure - does not apply while Trace mode is enabled. Deactivate Trace mode immediately after debugging.
How Message Processing Logs Reach the Cloud Cockpit

Activating Trace Mode for Detailed Message Analysis

For standard operations, MPLs show the outcome and error message. When you need to understand exactly what happened inside an iFlow - for example, to diagnose a mapping error - activate Trace mode.

  1. Navigate to Monitor → Manage Integration Content.
  2. Select the integration flow you want to trace.
  3. Click Log Level and set it to Trace.
  4. Send a test message to the iFlow endpoint.
  5. Open the resulting MPL and navigate to the Trace tab.
  6. Step through the iFlow execution: each step shows the message content (headers, payload) at that point.

Note

Trace mode generates significantly more data than standard logging. In EIC, trace data is stored in the local database (PostgreSQL or SAP HANA) and transmitted to the Cockpit, consuming both local storage and network bandwidth. Enable Trace mode only for the duration of your debugging session, then reset to the Info or Error level. The Trace activation and navigation experience is identical on EIC and the Cloud Integration Runtime - existing trace debugging skills transfer directly.

Investigating a Failed Message: A Structured Approach

When a message shows status Failed, start at the application layer before moving to infrastructure.

  1. Open the MPL. Click on the failed message entry in Monitor → Message Processing.
  2. Read the error message carefully. Common patterns: Connection refused / HTTP 503 → target system unreachable; Authentication failure → credentials incorrect or expired; Mapping error → missing required field in source; Timeout → target system did not respond within configured timeout.
  3. Activate Trace mode and reprocess a test message if the error is not immediately clear.
  4. Navigate through the trace to identify the exact step that failed and the message content at that point.
  5. If the error is application-level (iFlow logic, target system connectivity), resolve it at that layer. Only escalate to the platform team for Kubernetes-level investigation if application-layer evidence is absent.

Monitoring API Proxies on EIC

API proxies deployed to EIC are monitored via API Analytics (request volume, latency, error rates, policy violations) and API Health (deployment status, runtime errors) - both accessible through the Cockpit using the same Runtime filter that scopes Message Processing to your EIC instance.

MPL Persistence and Retention

MPLs in EIC are stored in two locations: locally in EIC's database (PostgreSQL or SAP HANA) (primary storage during execution and connectivity gaps) and in the SAP Integration Suite cloud monitoring store (for Cockpit access). For long-term archiving to meet compliance requirements, note that MPL archiving to external storage systems is currently a known EIC limitation. Always consult SAP Note 3391207 for the current status of this capability.

Beyond the Cockpit: SAP Cloud ALM and SAP Focused Run

Enterprise Monitoring: SAP Cloud ALM and SAP Focused Run

For organizations operating multiple EIC instances - or requiring centralized monitoring across a hybrid integration landscape - EIC integrates with two SAP enterprise monitoring platforms. Both preserve data residency: only operational metadata crosses the network boundary. The business message payload remains within your private infrastructure.

SAP Cloud ALM (Recommended)

SAP Cloud ALM is SAP's cloud-based Application Lifecycle Management platform. For EIC, it supports push-based transfer of MPL metadata from each Edge Node to the central Cloud ALM tenant, enabling a unified view of your entire integration landscape.

What SAP Cloud ALM provides for EIC:

  • All EIC runtimes across your landscape appear in a single Cloud ALM monitoring dashboard.
  • Alerts configurable based on MPL status: failed messages, processing anomalies, and high latency.
  • Only monitoring metadata (status, timestamps, error codes) is transferred - the payload stays in the private database.
  • During connectivity loss, MPLs from offline periods are transmitted to Cloud ALM once connectivity is restored.

Available monitoring categories in SAP Cloud ALM for EIC:

CategoryTypeDescription
Edge Integration MessagesMessagesAll messages processed by EIC instances
Failed MessagesEventFailed MPLs, individually or grouped
Anomalies in Processing VolumeEventUnusual message volume patterns
High Processing TimeEventMessages exceeding expected processing time

Setup requires connecting your SAP Integration Suite subscription to SAP Cloud ALM via the Landscape Management Service, then following the SAP Cloud ALM product-specific monitoring setup guide for Integration Suite.

SAP Focused Run

For organizations with SAP Focused Run (FRUN) as their primary ALM tool - common in on-premise-heavy landscapes - EIC is supported from FRUN 4.0 FP02+ (Health Monitoring) and FRUN 5.0 FP02+ (Integration and Exception Monitoring). The architecture follows the same push-based model: monitoring metadata is transmitted from EIC to Focused Run; business message payload remains local.

Note

The data-residency principle applies to both SAP Cloud ALM and SAP Focused Run integration: only operational metadata crosses the network boundary. This is a deliberate design decision, not a technical limitation.

Scenario Check-in: Global Manufacturing AG

Global Manufacturing AG's integration team monitors the S/4HANA ↔ QMS iFlow daily through the SAP Integration Suite Cockpit, using the Runtime filter to scope all views to FRA-EIC-PROD; MPLs are generated locally by the EIC runtime and transmitted asynchronously to the Cockpit - with quality inspection data never leaving the private infrastructure. When a second EIC instance is added at another facility, both appear in the same SAP Cloud ALM dashboard, giving the integration team enterprise-wide visibility without any business payload crossing the network boundary. At the infrastructure layer, the platform team monitors Kubernetes health, pod status, and resource utilization independently using Prometheus and Grafana - with neither team performing the other's role. This operational separation mirrors EIC's architectural principle: the integration team owns the application layer; the platform team owns the infrastructure layer.

Key Takeaways

In this lesson, you learned how the SAP Integration Suite Cockpit serves as the primary monitoring interface for integration flows running on EIC, with the Runtime filter as the key step for scoping all views to a specific EIC instance. Message Processing Logs (MPLs) provide the complete audit trail for every processed message - stored locally in EIC first and transmitted asynchronously to the Cockpit - and Trace mode enables step-by-step iFlow debugging with identical behavior to the cloud runtime. This application-layer monitoring is one half of a two-layer strategy: infrastructure-layer monitoring - pod health, resource utilization, cluster status - remains the platform team's responsibility, using Kubernetes-native tools such as Prometheus and Grafana. For organizations operating multiple EIC instances, SAP Cloud ALM and SAP Focused Run provide centralized visibility across the full landscape, transmitting only operational metadata - never the business payload.

Lesson 4.2 builds on this operational foundation to answer the planning question: what exactly can EIC do today, what are its current boundaries, and where is it headed?