n8n published [email protected] on October 26, 2026, with a core fix that starts a subworkflow’s error workflow when a check before execution fails. For operators who depend on error workflows to report failed ETL jobs, this is the main change. This release is not a prerelease and also addresses IMAP reconnects and SharePoint mapping.
The full release notes and downloads are on the GitHub release page.
Error workflows cover checks before execution ¶
The core correction in #41978 concerns a failure before a subworkflow starts running. The release fixes the error workflow startup path for that case. That timing matters when a pipeline delegates part of its work to another workflow and relies on the associated error workflow for reporting.
An error handler only helps if execution reaches it. If your handler sends an alert or records a failed ingestion attempt, this fix makes the check failure relevant to that reporting path. Those downstream actions still depend on what you configured in the error workflow.
A useful regression check is to reproduce a failing check before subworkflow execution, then confirm that its error workflow starts. Testing an error inside an already running subworkflow exercises a different point in execution. Include both cases when validating the behavior your pipeline expects.
IMAP reconnects back off after repeated drops ¶
The Email Trigger IMAP node now backs off reconnect attempts when the server drops each new connection, as described in #41927. The trigger condition is specific: each replacement connection gets dropped too.
For a mailbox ingestion pipeline, repeated reconnects create work without restoring message intake. Backoff should reduce that connection churn while the server continues dropping connections. The tradeoff is time between attempts, which belongs in any assessment of how quickly the trigger recovers after the server becomes usable again.
The notes do not specify the delay schedule or a tuning option. When checking this fix, observe reconnect timing during repeated drops and verify that ingestion resumes after connections remain open. A single disconnect followed by a successful reconnect does not reproduce the failure pattern described here.
SharePoint automatic mapping fixes two symptoms ¶
The Microsoft SharePoint node correction in #41906 covers a column mapper 403 and empty items in Map Automatically. These are relevant to workflows that depend on SharePoint field mappings to move data between systems.
The practical check has two parts. Confirm that the column mapper works for the affected setup, then inspect the items produced with Map Automatically. A mapper that opens successfully only verifies one of the reported symptoms. Compare the mapped values against the source data before accepting the result.
Keep the scope of the 403 fix precise. The release identifies a column mapper problem; it does not describe a general change to SharePoint access permissions. Existing authorization requirements still need to be evaluated separately from the mapper behavior.
Task runner certifi update ¶
The release also updates the task runner’s certifi dependency, with the later listed update targeting 2026.10.23 in #41994. This is dependency maintenance in the runner and belongs in the runtime change record for deployments that use it.
The notes give a package version but name no affected endpoint, certificate, or security advisory. That supports recording the dependency update; it does not establish a specific incident or vulnerability that this release resolves.