SAP Message Monitoring in CPI
An integration flow that runs silently is not necessarily running correctly. A message can fail partway through processing, a receiver system can be unavailable, or a mapping can produce wrong output — all without anyone noticing unless monitoring is in place. Message monitoring gives you visibility into every message that passes through CPI, so you catch problems before they affect the business.
Think of message monitoring like the control room at an airport. Every flight appears on the screen. Controllers see which flights are on schedule, which are delayed, and which have problems. Without the control room, planes fly and nobody knows what is happening. CPI monitoring is your integration control room.
The Monitor Section in CPI
Access monitoring from the SAP Integration Suite home page by clicking Monitor. The Monitor section has several subsections:
- Message Processing – View the processing status of every message that ran through deployed iFlows
- Integration Content – View and manage deployed iFlows and their runtime status
- Manage Security – Manage credentials, certificates, and keystores
- Manage Stores – View Data Store entries and message queues
- Manage Locks – View and release idempotent locks
Message Processing Log (MPL)
Every message processed by a deployed iFlow generates a Message Processing Log (MPL). The MPL is the primary diagnostic tool for integration developers. It records the complete lifecycle of a single message from arrival to delivery.
MPL Status Values
COMPLETED ✓ Message processed successfully end to end FAILED ✗ An error occurred; processing stopped RETRY ↺ Processing failed and is being retried PROCESSING ⟳ Message is currently being processed (in flight) DISCARDED ○ Message skipped (e.g., duplicate detection triggered) CANCELLED ✕ Message was manually cancelled
Reading an MPL in Detail
Clicking an MPL entry opens the detail view. It shows:
- Overview – Start time, end time, duration, overall status, iFlow name
- Logs – Step-by-step processing trace showing each iFlow step and its outcome
- Properties – Custom log entries your iFlow added using the Script step
- Attachments – Message payloads saved at specific points using the Trace or Persist step
Filtering Messages in Monitor
Production CPI tenants process thousands of messages daily. The monitor provides filters to find specific messages quickly:
Filter options: Time period: Last hour / Last day / Last week / Custom range Status: All / Completed / Failed / Retry iFlow name: Search by integration artifact name Correlation ID: Find all messages belonging to one business transaction Custom status: Developer-defined status labels
Set a narrow time range and specific iFlow name first. This returns a manageable number of results instead of thousands of entries.
Correlation IDs
A correlation ID links multiple messages that belong to the same business transaction. For example, a sales order integration might trigger three separate iFlows — one to create the order, one to check inventory, and one to send a confirmation email. Using the same correlation ID in all three lets you find all related messages in one monitor search.
Set the correlation ID in a Content Modifier early in the iFlow:
Message Header:
Name: SAP_ApplicationID
Value: ${property.SalesOrderNumber}
The monitor displays this value in the Application Message ID column.
Search the monitor by this ID to find all messages from one order.
Trace Mode for Debugging
The standard MPL shows high-level step outcomes. When debugging a complex problem, you need to see the actual message payload at every step. Trace mode captures and stores the payload before and after every processing step.
To activate trace mode:
- Navigate to Monitor → Manage Integration Content
- Find the deployed iFlow
- Click the iFlow and set Log Level to Trace
- Reproduce the problem by sending a test message
- Find the resulting MPL and click through each step to see the payload at that point
- Set Log Level back to Info after debugging — trace mode generates large amounts of data
Trace mode is only for temporary debugging. Running it continuously in production fills storage quickly and slows performance.
Persisting Payloads for Audit
For compliance or audit purposes, you may need to store copies of processed messages permanently. Use the Persist step (called "Write Variables" in newer CPI versions) to save payload snapshots to the Data Store. Unlike trace mode, which is temporary, Data Store entries persist until explicitly deleted or expired.
Monitoring Deployed Artifacts
Beyond individual messages, the Integration Content monitor shows the health of all deployed iFlows:
STATUS INDICATORS: Started (green) – iFlow is deployed and processing normally Error (red) – iFlow failed to start or crashed during startup Stopping – iFlow is in the process of being undeployed Stopped – iFlow is undeployed and not processing messages
An iFlow in Error state at the artifact level means no messages are being processed at all — typically caused by a configuration error (wrong credentials alias, invalid XSD, missing required parameter). Fix the configuration and redeploy.
Setting Up Operational Alerts
Manual monitoring requires someone to check the monitor screen regularly. Automated alerting notifies the team proactively when something goes wrong. CPI provides alert rules you can configure:
- Alert when a specific iFlow has more than N failed messages in the last hour
- Alert when an iFlow's average processing time exceeds a threshold
- Alert when a deployed iFlow enters Error status
Alert notifications go to email addresses you configure. Connect them to your team's ticketing system or incident management platform for a complete operational loop.
Data Volume and Retention
CPI stores MPLs for a configurable retention period. The default varies by plan. After the retention period, old MPLs are deleted automatically. If your business requires longer message history — for example, financial audit requirements of seven years — export MPL data to a separate log management system using CPI's OData API for monitoring data, or the Audit Log integration pattern.
Monitoring Checklist for Operations Teams
- Check the Failed message count every morning before business hours start
- Investigate any iFlow showing Error status immediately
- Review processing duration trends weekly — sudden increases indicate performance problems
- Confirm all critical iFlows processed their expected message volumes (no messages = possible upstream problem)
- Review retry queues for stuck messages that have been retrying without success