What Happens If My Import Fails Halfway Through?

AccuArk uses batch processing with database transactions to ensure your data remains consistent, even if something goes wrong during an import.

How Batch Processing Works

Rather than processing your entire file as a single operation, AccuArk divides the import into batches of 100 rows. Each batch is wrapped in a database transaction:

• If all rows in a batch succeed, the entire batch is committed (saved permanently)
• If any row in a batch fails, the entire batch is rolled back (no records from that batch are saved)

This means you will never end up with a partially imported batch. Either all 100 rows in a batch are saved, or none of them are.

What Happens to Successful Batches

Batches that have already been committed are not affected by failures in later batches. For example, if your file has 500 rows and the third batch (rows 201 through 300) encounters an error:

• Batch 1 (rows 1 through 100) — committed successfully
• Batch 2 (rows 101 through 200) — committed successfully
• Batch 3 (rows 201 through 300) — rolled back due to error
• Batch 4 (rows 301 through 400) — continues processing
• Batch 5 (rows 401 through 500) — continues processing

The data from batches 1, 2, 4, and 5 remains in the system. Only the failed batch is rolled back.

Getting an Error Report

After the import completes, the summary screen shows exactly how many rows succeeded and how many failed. You can click Export Error Report to download a CSV file that contains:

• Each row that failed
• The specific error message for that row
• The original data from your file

Use this report to identify and fix the problem rows.

Re-Importing Failed Rows

After fixing the errors, you can re-import just the failed rows. For data types that support upsert (such as Inventory Items, Customers, and Vendors), rows that were already imported successfully will simply be matched and updated — you will not get duplicates. This makes it safe to re-import a corrected version of your original file.

For data types that do not support upsert (such as Outstanding Invoices or Bank Transactions), import only the corrected error rows to avoid creating duplicate records.

Summary

• Imports are processed in batches of 100 rows
• A failed row causes its entire batch to roll back, but does not affect other batches
• Successfully committed batches are permanent
• The error report tells you exactly which rows failed and why
• For upsert-supported data types, re-importing is safe and will not create duplicates