AEP failed batch and partial ingestion analysis in one click [Lens]

ADOBE EXPERIENCE PLATFORM (AEP)

Pradeep Jaiswal

10/9/20265 min read

Chrome extension. Link

What Is Lens?

Lens is a Chrome extension built for teams running Adobe Experience Platform. It works inside the AEP interface you already have open and uses your existing session: no extra login, no API credentials to configure, no data export step. Earlier articles covered the Schema ERD Canvas, the Source Dataflow Canvas, the Approved Component Highlighter and, in Part 1, Analyze Failed Batch for a single batch. This article covers the lists around it.

Failed Batch List: Everything That Broke, in One Table

Instead of opening dataflow after dataflow looking for a red badge, the Failed batch list shows the newest 1,000 failed batches from the last 90 days.

A few details make it practical for triage:

- Sort by Records Failed count
- Sorting by Error groups failures by code
- Filter as you type across the columns

Click a Batch ID hyperlink and you are in the single-batch analysis sub tab, with nothing to copy and paste.

Try Lens

Who It’s For

Data engineers get a standing view of what is failing and what is silently losing records, without walking APIs by hand.

Implementation consultants can audit a client sandbox live, from the page the client already has open: which flows are unprotected, which batches failed, which runs lost rows.

Support and operations teams can answer “what failed since Tuesday?” from one table, then attach the exact rows to the ticket.

Data owners can see whether a failure is a one-off or the same code repeating across runs, and fix it at the source.

Practical Tips

Check the Dataflow list first. A flow listed there will give you a code and a count and nothing more on its next failure.
Mind the windows. The lists look back 90 days, and Adobe deletes failed-file and diagnostic files after 30 days, so open a batch while its files still exist.
Treat partial ingestion as a signal, not a pass. A green run with a nonzero Records Failed is a data-quality issue waiting for a report to expose it.
Press Show more only when you need it. The preview is usually enough to name the cause.

Part 1 helps once you have a batch in front of you. Part 2 is about not needing someone to hand you one: finding the unprotected dataflows, the failed batches and the quiet partial failures from the page you are already on.

Keep Exploring Lens

Start with [From “Failed” to “Fixed”: How Lens Shows You Exactly Which Records Broke Your AEP Ingestion], which covers the Single-batch analysis in depth. To see where the data comes from, read Follow the Data [How Lens Maps Every Source Flow Across Your Adobe Experience Platform]. To see how your schemas relate, read Seeing the Invisible [How Lens Makes Your Adobe Experience Platform Schema Architecture Visual]. And if your team reports in Customer Journey Analytics or Adobe Analytics, read Every Analytics Team Has Been Here [How Lens Makes Governance Visible Inside CJA and Adobe Analytics].

Part 1 of this series covered what happens once you are looking at one failed batch: the batch error, the rows that broke and the file they came from. But there is also a quieter version of the same problem: runs that Adobe marks Success while quietly rejecting some of the rows, and dataflows that are set up so that, when they do fail, Adobe will not keep the per-record reasons at all.

The Analyze Failed Batch left rail in Lens now has four sub-tabs that cover that whole picture:

Dataflow List: Fix It Before It Fails

When a batch fails, the most useful thing Adobe can keep is the per-record error file: every rejected record with its error code, the column and the message. Adobe only keeps it if Error diagnostics was switched on for the dataflow. If it was off, you get a count and a code, and nothing more.

The Dataflow list shows you the dataflows where that protection is missing. It lists your source dataflows that have Error diagnostics disabled, or partial ingestion off. It shows the newest 1,000 such dataflows. Audience and system dataflows, and datasets Adobe creates itself, are left out, using the same rule as the Source Dataflow left rail, so the list contains the flows you actually own.

Partial Ingestion List: The Failures That Say “Success”

Partial ingestion is a setting that lets a batch succeed even though some records were rejected, up to a threshold you choose. It is useful, and it is also how data goes missing without any red badge. The run is green, the dataset has rows, and a slice of the source never arrived.

The Partial ingestion list has the same shape as the Failed batch list, plus one extra column, Records Ingested, because here the number that landed matters as much as the number that did not. The status reads Success (partial ingestion).

Read records Ingested next to records Failed on the same row to judge how much was lost, then open the batch to see which records and why.

Batch ID Details: From a Row in a List to the Failed Records

Every list row from Lens UI , Lens logo icon or ‘Analyze in Lens’ cta from AEP UI opens the same detail view in Lens. If you arrive from a Failed run you see the batch status, the dataset and dataflow, the record counts and the batch-level errors. Below that sit the two views of the failed data:

- Row Error file: every failed record with its error code, failing column and message, plus the source file it came from.
- Batch file: the original input file that was rejected, before any mapping.

The behaviours matter on large batches: A quick preview first, the whole file on demand. The first records appear within seconds, read from the start of the file. When you need more, Show more loads the whole file, up to 10,000 rows in the table, and every record goes into an export. Files too large to hand over in one piece are transferred in smaller pieces, so a very large error file loads instead of failing.

Need Expert Help Fixing Adobe Experience Platform Ingestion Failures?

At Shiftlytic, we specialize in Adobe MarTech solutions, helping brands like yours find why batches fail, fix data quality issues at the source, and keep Adobe Experience Platform pipelines healthy. Whether you're stuck on recurring failed batches, unclear dataflow errors, XDM mapping and schema mismatches, or a source connector that keeps dropping records, our team of Adobe-certified experts is here to guide you.

Don't let failed ingestion leave gaps in your profiles and reports. Contact us today to schedule one free consultation and get your data flowing reliably, accurately, and ready for activation. Let's turn your data pipeline into a competitive advantage together.