# Automation troubleshooting

> Fix disabled automations, failed steps, and other common Automation Studio errors.

## Overview

Use this page when an automation does not trigger, triggers too often, or a developer step fails. Each section lists the usual reasons, then how to resolve them.

> **Note:**
> Email, Slack, Teams, surveys, Google Calendar, and document steps are covered in [Automation Studio troubleshooting guide](https://help.rippling.com/article?articleId=25bWg4rdjPLI2xKD19cVFe) in the Help Center.

* * *

## An automation was automatically disabled

If an automation encounters an error more than 10 consecutive times and the failures persist for over 24 hours, the automation will be automatically deactivated. You will receive a notification.

This can occur when an automation triggers but fails to complete a step. For example, if an automation triggers but it’s unable to complete an email step because the group for the recipients is empty, it can fail repeatedly, causing this to occur.

> **Note:**
> If an automation fails fewer than 10 times or if it fails but then succeeds again within a short period of time, it will remain active. Automations are only deactivated when they are persistently encountering errors for a long time.

Automations that were automatically deactivated due to failures will appear as **Inactive** on the automations dashboard.

You’ll need to re-activate the automation before it will run again. Before you re-activate it, review its settings to check for any misconfigurations that are causing it to fail.

To see more details about the failure:

1.  From the sidebar, select **Tools** > **Automation Studio**.
2.  On the **Dashboards** tab, open the automation that is failing.
3.  Select the **Execution history** tab.
4.  For rows where the **Status** column says _Failed_, review the **Failed step** column to see exactly where the failure occurred.
5.  Click **Edit automation**, find the failing step, and review its settings.

When you’re done, save any changes and re-activate the automation.

A failed later step does not undo an earlier one. Once a step succeeds, its changes are committed, including a custom object write or a function that already ran. For more information, see [Automation steps](https://developer.rippling.com/documentation/rippling-platform/developer/workflows/steps#using-data-write-steps).

* * *

## Unable to set up, save, or activate an automation

If you’re unable to set up, save, or activate an automation, it may be for the following reasons:

-   Subscription level does not allow API usage
-   Configuration is incomplete
-   Lack of permissions

### Subscription level does not allow API usage

Only customers with a subscription to the Platform Pro tier can use the **Call a public API** step. Customers on the Core tier or legacy Pro tiers do not have access to this step. If you don’t have access, contact your Account Manager to request transfer to the new plan. For more information, see [Call an API or webhook using an automation](https://developer.rippling.com/documentation/rippling-platform/developer/workflows/call-api).

### Configuration is incomplete

You cannot publish an automation until all its steps have been fully configured. Steps that haven’t yet been fully configured show “More information required” on the automation canvas.

To resolve this issue, click on each step and fill in all of the required fields. Required fields are indicated with an asterisk.

### Lack of permissions

You may not have permissions to view or use specific fields in the trigger or conditions, or to configure certain steps. Super and full admins have permission to use all available automation configuration options. Other admins may not be able to access certain fields based on their permissions. For more information, see [Automation permissions](https://developer.rippling.com/documentation/rippling-platform/developer/workflows/permissions).

* * *

## Automation is not triggering

If your automation is not triggering when you expect it to, it may be for the following reasons:

-   The automation was inactive when it was supposed to trigger
-   Multiple versions of the automation are running
-   The owner of the automation has limited permissions
-   The automation uses _Employee start date_ instead of _Start date_
-   The data the automation uses in the trigger was not available at the time it triggered
-   Mutually exclusive conditions
-   Time zone difference with the automation owner
-   The automation uses a set schedule and was updated
-   Time of day misalignment for a trigger relative to a date
-   The trigger uses _Employee is updated_ with the _Teams_ field as a condition
-   The automation uses a one-time schedule set to a boundary minute

### The automation was inactive when it was supposed to trigger

Automations only trigger when they are published and active. Draft automations and those in an inactive state do not trigger. Check the status of the automation at the time you expected it to trigger.

1.  On the Automation Studio dashboard, find the automation and click the arrow at the end of the row to open its details.
2.  Select the **Activity log** tab.
3.  Identify whether the automation has been deactivated before the expected trigger. The activity log will have a row named “Updated automation status” whenever the automation is activated or deactivated.
4.  Check whether the automation was ever activated or temporarily deactivated by another team member when it was expected to trigger.

For more information, see [Automation statuses](https://developer.rippling.com/documentation/rippling-platform/developer/workflows/statuses).

### Multiple versions of the automation are running

Automations can include pauses and delay steps that cause them to run over a long period of time. When you update an automation to change how it behaves, a new version is created. Any automations that are still actively running in the background will operate according to the configuration in the prior version of the automation. For example:

-   You publish V1 of an automation. It includes a 7-day delay.
-   You then update the automation so it becomes V2.

Any automations that triggered after you published V1 and before you published V2 will continue to run in the background for 7 days, completing all steps according to the configuration in V1. Any automations that trigger based on V2 of the automation will behave according to the configuration in V2, which may differ from V1.

To verify the expected behavior, cross-check the trigger history with the published version history.

### The owner of the automation has limited permissions

The owner of the automation must have permissions to view the data used in the automation for the target employees. Check the information used in the trigger and the people included in subsequent steps, and confirm the owner has sufficient permissions. For more information, see [Automation permissions](https://developer.rippling.com/documentation/rippling-platform/developer/workflows/permissions).

If the automation trigger uses the _Employment status_ field and includes workers that are not active (new hires or terminated), the automation will only trigger if the owner has permission to hire and terminate.

### The automation uses _Employee start date_ instead of _Start date_

Rippling has two similarly named fields that represent different dates:

-   _Start date_. This is the worker’s start date of their current employment period with your company.
-   _Employee start date_. This is the date a contractor’s employment type transitioned to employee.

For most common use cases, you’ll want to use _Start date_. Double-check that your automation uses the correct date.

### The data the automation uses in the trigger was not available at the time it triggered

An automation may not trigger if one of its filter conditions relies on data that is not yet populated.

For example, if an automation triggers when an offer is accepted but it has a condition that relies on home address, it may not trigger. This is because typically the home address field doesn’t get populated until later in the onboarding process.

To see whether this is what caused your automation not to trigger, review the fields used in the trigger filter conditions and ensure they are available at the moment the trigger event occurred. You can cross-check the automation trigger history with the employee’s change history to see whether a field was populated before or after the automation triggered.

### Mutually exclusive conditions

If there are mutually exclusive conditions on a trigger (meaning _AND_ conditions), make sure that those conditions are logically possible. For example, suppose you have a condition for “Department is A” AND “Department is B”, but an employee can only be in A or B. In this case, the mutually exclusive condition will never be met and therefore the automation will not trigger.

To resolve this issue, review and update any complex AND conditions in your automation.

### Time zone difference with the automation owner

Scheduled triggers and date-relative triggers operate based on the automation owner's local time zone.

If you are not the automation owner and the automation is not triggering at the time you would expect, check whether your time zone in Rippling is different from the time zone of the automation owner.

To check the time zone of the automation owner, you will need to reach out to them directly. Rippling users can only see their own time zone within their account settings.

### The automation uses a set schedule and was updated

When an automation uses the trigger **On a set schedule**, any changes made to the automation after it has already run will cause the schedule to reset, affecting the next execution. For example, if an automation is scheduled to run biweekly every Monday and it successfully ran on Monday, Jan 26, but the name of the automation is modified on Tuesday, Jan 27, the automation will **not** run on the upcoming Monday, Feb 2. Instead, it will be deferred and run on the next scheduled Monday, February 16.

If you continue to update the automation frequently, the first occurrence will continue to be pushed back, which can prevent the automation from triggering at all. This behavior applies to all frequency options.

### Time of day misalignment for a trigger relative to a date

If the automation uses the trigger **Relative to a date**, there may be a timing conflict between the scheduled automation run time and the effective time of the triggering change event.

When you make an update to the employee profile, you're prompted to select the specific date and time for the change to become effective. If the effective time for the change occurs after the scheduled run time for the automation, the automation will not trigger.

For example, say you have an automation that is set to trigger **Relative to a date** > **Start date**.

-   The automation is scheduled to run at 11:00 AM Pacific Time.
-   You update a new hire's start date and set the effective time of the change to be 12:00 PM Pacific Time.
-   The employee's new start date becomes effective one hour after the automation was scheduled to run. Therefore, the automation will not trigger.

To ensure automations trigger correctly, ensure that the scheduled automation run time occurs after the effective time for the change.

### The trigger uses _Employee is updated_ with the _Teams_ field as a condition

When using _Employee is updated_ as the trigger and the _Teams_ field as a condition, the automation will only trigger when configured in a specific way. For more information, see [Automation triggers](https://developer.rippling.com/documentation/rippling-platform/developer/workflows/triggers#object-creation-and-updates).

### The automation uses a one-time schedule set to a boundary minute

Scheduled automations are evaluated by a background job that runs every 15 minutes, at :00, :15, :30, and :45. For one-time schedules, each sweep checks a 14-minute window where the scheduled time must be greater than or equal to the window start and strictly less than the window end.

This means times that fall on the last minute of a window, specifically minutes ending in :14, :29, :44, or :59, are not picked up by any sweep. For example, an automation scheduled at 6:59 AM will not match the 6:45 AM sweep (because 6:59 is not strictly less than 6:59) and will not match the 7:00 AM sweep (because 6:59 is before 7:00).

To avoid this issue:

-   Do not set one-time scheduled automations to times ending in :14, :29, :44, or :59.
-   Use exact 15-minute boundaries (:00, :15, :30, or :45) when possible.
-   If the automation must run at a specific time, round to the nearest safe minute. For example, use 7:00 AM instead of 6:59 AM.

Note that scheduled automations may not execute at the exact second shown in the UI. Execution is matched during the 15-minute background sweep and can be slightly staggered for load distribution.

* * *

## Automation is triggering unexpectedly or too often

If your automation is triggering unexpectedly, it may be for the following reasons:

-   The trigger condition is too broad
-   _Effective date_ is used instead of _change date_, or vice versa

### The trigger condition is too broad

If your automation is using a trigger condition like "Profile change is effective" or "Employee is updated," and no narrowing filters have been applied to that condition, then the automation will run for any update, including system updates or formula field recalculations.

To resolve this issue:

-   If the trigger condition is a generic "Object X is updated," identify the specific change you want to capture and add a condition to the trigger.
-   If the trigger condition has a filter, check that the filter is not “Field is X” if you intended for the automation to only trigger when the field is _changed_. If the filter is "Field is X", then update the filter to be "Field changes to X".
-   If the trigger condition does not have any filters, add filter conditions such as "Field X changes to Y" or "Field Z is updated".

**Changes** and **Changes to** apply only when the selected field changes. **Before the change** and **After the change** apply regardless of which field changed. For more information, see [Filter conditions](https://developer.rippling.com/documentation/rippling-platform/developer/workflows/filter-conditions#add-conditions-that-trigger-based-on-a-record-change).

### _Effective date_ is used instead of _change date_, or vice versa

When a change is made in Rippling, you can typically choose the _effective date_ of the change. This may or may not be the same as the _change date_, i.e., the date the actual change was made in Rippling. If your automation is triggering on an unexpected date, check to see if the automation is configured to trigger on the effective date of the change or on the change date.

* * *

## A developer step fails

If the automation triggers but a query, function, API call, bulk operation, delay, or SFTP step fails, check **Execution history** for the failed step, then use the causes below.

For more information, see [Automation limits](https://developer.rippling.com/documentation/rippling-platform/developer/workflows/limits) and [Automation steps](https://developer.rippling.com/documentation/rippling-platform/developer/workflows/steps).

### The step hit a limit

-   **Query Rippling data** can retrieve a maximum of 50,000 records in one automation run. If a collection contains more than 50,000 records, the automation may fail.
-   **Bulk operations** cannot iterate more than 1,000 times.
-   A **function** step times out after 50 seconds. Use a separate function step for each operation. Loop items can include trailing whitespace. In the function, call `.trim()` on the loop item before comparing it to an API response. For more information, see [Automation steps](https://developer.rippling.com/documentation/rippling-platform/developer/workflows/steps#bulk-operations).
-   An automation cannot run for more than 45 days total. A single delay step cannot exceed 30 days.
-   Calls to external applications, including APIs and webhooks, time out after 60 seconds.
-   Scheduled automations with filter conditions run once for each matching record, up to 5,000 records. If there are more than 5,000 matching records, or if no records match the condition, the automation will not run.

### A delay step is using stale trigger data

When an automation is triggered, the data related to the trigger event is available to use in downstream steps. This data is generated at the time of the trigger event, and is not updated again later, even if the automation includes delays. If your automation includes a delay and you want to refresh the data in the automation, add a step to query Rippling data.

For example, say you have an automation that sends reminders for unsigned documents. You want the automation to send an initial notification as well as a follow-up reminder after 7 days. After the 7-day delay step, you’ll need to add a step to query the status of the documents. This way, the automation will have the latest status and can send reminders only for documents that remain unsigned.

* * *

## Call a public API fails

An automation with a step that calls a public API may fail for one of the following reasons:

-   Incorrect HTTP method (GET, POST, PUT) for the endpoint.
-   Incorrect authentication method.
-   Invalid authentication token or key.
-   Incorrect URL (either typos or space in URL).
-   Payload formatting: Incorrect JSON structure, wrong data types, missing required fields, incorrect date formats, values not accepted by the destination API.
-   Timeout: API steps fail if Rippling doesn’t receive a response from the third-party API after 60 seconds.

To resolve these issues:

-   Check the third-party API documentation for the correct HTTP method and authentication to use for the endpoint.
-   Check the third-party API documentation for the correct expected format for each input in the payload (for instance, should dates be formatted mm/dd/yyyy or dd/mm/yyyy?).
-   Check the endpoint URL for any typos or issues.
-   Test the endpoint and authentication key outside of Rippling in an API testing environment (Postman, SoapUI, or equivalent) to isolate the issue.

For more information, see [Call an API or webhook using an automation](https://developer.rippling.com/documentation/rippling-platform/developer/workflows/call-api).

### Can I use `\n` to create line breaks in a Slack channel message or an API webhook payload?

No. The `\n` or `/n` characters for line breaks are not supported when sending messages to Slack channels or using the **Call a public API** step in the Automation Studio app. When an automation triggers these steps, Slack or the third-party API server interprets the characters as plain text rather than a line break.

While line breaks are supported when sending direct messages (DMs), Rippling cannot control how third-party platforms or external servers interpret specific markup in a channel message or a JSON payload. Using alternative characters, such as //n, will also result in the characters displaying as plain text.

This issue typically occurs when you attempt to compile and send a dynamic list, such as a list of employee names, to a public Slack channel. Because the system passes the data as a continuous string to the external channel, standard code formatting characters like `\n` fail to render as formatting commands.

* * *

## Send a report via SFTP fails

An automation with a step that sends a report via SFTP may fail for one of the following reasons:

-   Incorrect authentication method: Incorrect username, password, SSH key, or port. For example, the SFTP server might require a public key auth, but the step is trying a password.
-   Lacking permissions on the host server: the SFTP user lacks write permissions or permission to create a directory on the SFTP server.
-   File path: Invalid destination file path (for example, a space in the path provided).
-   Report access: The report is sent using the automation owner's permissions. Ensure the automation owner still has access to the report.
-   Report type: Some special report recipes cannot be sent via an automation.
-   Server-side issues: For example, SFTP server overload, network instability on the server's end, or rate limiting.

To resolve these issues:

-   Run a basic test of your SFTP credentials in your command terminal. You can type `sftp your_username@sftp.example.com`. Replace `your_username` and `sftp.example.com` with your credentials and server.
-   Confirm the SFTP user has permission to create files and folders on the SFTP host. Users who cannot create new folders will see failures.
-   Confirm the report is not one of the special report recipes that automations cannot send. For more information, see [Automation steps](https://developer.rippling.com/documentation/rippling-platform/developer/workflows/steps#send-a-report-via-sftp).
-   Confirm the automation owner has permission to access the report.

* * *

## See also

-   [Get started with automations](https://developer.rippling.com/documentation/rippling-platform/developer/workflows/get-started)
-   [Automation permissions](https://developer.rippling.com/documentation/rippling-platform/developer/workflows/permissions)
-   [Automation triggers](https://developer.rippling.com/documentation/rippling-platform/developer/workflows/triggers)
-   [Filter conditions](https://developer.rippling.com/documentation/rippling-platform/developer/workflows/filter-conditions)
-   [Automation steps](https://developer.rippling.com/documentation/rippling-platform/developer/workflows/steps)
-   [Call an API or webhook using an automation](https://developer.rippling.com/documentation/rippling-platform/developer/workflows/call-api)
-   [Automation statuses](https://developer.rippling.com/documentation/rippling-platform/developer/workflows/statuses)
-   [Automation limits](https://developer.rippling.com/documentation/rippling-platform/developer/workflows/limits)
-   [Invoke functions](https://developer.rippling.com/documentation/rippling-platform/developer/functions/invoking-functions)

Source: https://developer.rippling.com/documentation/rippling-platform/developer/workflows/troubleshooting

> For the complete documentation index, see [llms.txt](https://developer.rippling.com/llms.txt). Public pages have a Markdown version at the same URL with .md added.
