Common pipeline errors

You may occasionally see errors in pipelines. Use this article to help you troubleshoot some common issues.

Approaches to troubleshooting errors

This section discusses where can you find out about errors and what you can do about them.

Two common types of errors

Here are the common types of errors in pipelines:

Design-time errors

At design-time, pipeline reviews the following:

  • The type of data
  • The context/scope of the data

Problems in these areas can generate an error.

Runtime errors

When your pipeline runs, you may see:

  • Temporary, or transient errors
  • Errors returned from remote systems

Where errors display

You can see errors when you are creating a pipeline (a Design-time error) or when a pipeline runs (a Runtime error).

Design-time errors

In this example, we have an error on a Prepare Bulk Record Upsert:

Because the merge field was of the wrong type, this error message displayed:

Runtime errors

In this example, we don't see an error until the pipeline runs. We have entered a bad value for the SQL parameter:

and when we ran the pipeline, this error displays:

How to find logs

To view the activity log for a pipeline, click on the three dots next to the pipeline's name on the My pipelines page:

and click on View activity. The log for that pipeline displays:

How to read logs

If your pipeline has run, you will see details about when this pipeline ran and information about what happened. In our example, we could see that an error occurred (Validation error: Invalid date/time value) and details about the input and output that lead up to this error:

You'll notice that some messages in the log are in color. We use colors to indicate which data type is involved:

  • red = boolean
  • blue = numeric
  • green = string
  • black = object/null/undefined

In the examples that follow, you will see how these logs can be used to troubleshoot your error.

The most common errors

The following section discusses some common errors and best practices to troubleshoot them.

Third-party errors

Third-party errors can be the easiest type of errors to identify. They usually begin with Remote service… or Remote System…. You can search for these errors and the remote system’s API guide.

This is an example of an error that could be resolved by searching for the error code.

Text displayed in this error: Remote service reported an error: HTTP status code 404 for URL

In this example, we see an error that would likely need to be resolved in Sharepoint, rather than pipelines.

Text displayed in this error: Waiting for an Item Created event in Microsoft SharePoint...

Remote API errors

These errors come from requests built in either the JSON handler or Webhook channels and they can be difficult to troubleshoot. Sometimes these errors are straightforward, as in the example below. For more information about API errors, see the API guide.

Text displayed in this error: Remote API returned authentication error: Invalid API key. Please see http://openweathermap.org/faq#error401 for more info.

In this example, a pipeline is using an uncommon encoding type:

Text displayed in this error: {"error":"invalid_client_id","error_description":"Client None is not registered in the system."}"

Poller errors

These errors can occur for a variety of reasons, but you can use a splunk search tool to help you troubleshoot. In the first example is from the Sharepoint poller. Your first step for such an error would be to reconnect to Sharepoint in the pipeline editor.

Remote service reported an error: HTTP status code 404 for URL

This error is from data in a google sheet spreadsheet to which a pipeline was pointed. In this error there was an issue in the data on the sheet.

Text displayed in this error: Remote service reported an error: Invalid requests[999]

Invalid pipeline errors

You may see these errors happen when there is either a conflict or corruption in your pipeline’s YAML. This is often caused if you have edited your pipeline in multiple browser windows or if you deleted a large branch of a pipeline insted of deleting each individual step within your branch. You can avoid this by deleting each step from the bottom of the fork up rather than deleting a “for each loop” or “if branch”. These errors usually can be resolved by copying the pipeline or reimporting the YAML. In a worst case scenario, you may need to download the YAML from the last successful run as a guide and rebuild the pipeline from scratch.

Text displayed in this error: Invalid Pipeline: Inconsistency detected.

YAML import errors

These errors arre caused by a user error on import or because of an issue for a specific type of pipeline structure. Don't copy and paste the YAML from a non-plain text editor and verify that the you are using the correct account slugs: the placeholder for the user token/login credentials for that step.

This error below was due to the account slugs not being replaced with the correct ones in a new realm.

Text displayed in this error: Remote API returned authentication error

Unattached errors

Some pipeline errors and warnings aren't associated with any particular run ID and that may or may not appear in the run details. This usually indicates an underlying issue such as: memory issues in a Google Cloud Platform (GCP), a task that didn’t complete properly, etc.

This error happens when a ResumePoint isn’t stored when a pipeline run is moved from one task to another in Google cloud. The pipeline aborts since it doesn’t know where to resume in the run on the new task. In a case like that, you will see an error like this one:

Text displayed in this error: Internal error. Failed to resume execution.

If you see a pipeline error that references GCP like this one, open a support ticket.

False error

This error simply means that the pipeline was hitting the endpoint in Quickbase too quickly and the pipeline slowed down it's execution. These errors aren’t anything to be worried about until the retry is 10/15 or higher. However, this may be an indicator that your pipeline should be optimized for better performance.

Text displayed in this error: Quickbase app is experiencing unusually heavy traffic. Got a transient error from remote system. Retrying (1/15 with 7000ms timeout) with a back-off.

Alert

A pipeline may sometimes be throttled to slow down its execution. This is not an error or failure so much as a warning that the pipeline was hitting a Quickbase table too quickly.

Text displayed in this error: Throttling: Pipeline is delayed for 60 s

Jinja template errors

These errors usually mean that there is something incorrect in your jinja syntax. However, they can be related to more complicated issues such as underlying data, null values, data type mismatches, etc. This error is caused by spaces and line breaks in the jinja template and was resolved once the spaces and line breaks were removed.

Text displayed in this error: Validation error: Incorrect template

In this example we see a misleading data type error that isn’t the true source of the issue. When in doubt, try reformatting and making minor adjustments in your testing of a jinja expression. Also, the help topic About Jinja for more information on testing jinja/jinja syntax tips.

Text displayed in this error: Validation error: Incorrect template

Restful API errors

These error message can be anything from generic 500 failure errors to something obvious like an authentication error. For more information about API errors, see the API guide.

This error was caused by an extra space after QB-USER-TOKEN XXX-XXXX in the authorization header.

Text displayed in this error: Validation error: 401 Client Error: Unauthorized for url: https://api.quickbase.com/v1/records

For help troubleshooting an error like this, see the Errors page in the Quickbase Developer Documentation.

Text displayed in this error: Validation error: 502 Server Error: Bad Gateway for url:

Deleted pipeline

This is an error in Quickbase, but it specifically relates to a deleted pipeline that left behind an orphaned webhook in Quickbase. These errors are typically 404 errors. Verify that the webhook is orphaned and then delete it. A deleted webhook from a pipeline is automatically recreated when the pipeline is turned off and on again in case you inadvertently delete a webhook for an active pipeline.

Text displayed in this error: Not Found The requested URL was not found on the server. If you entered the URL manually please check your spelling and try again.

CSV data error

This error can occur almost anywhere in a pipeline from Quickbase channels to csv handler to onedrive. The error does tend to occur for specific reasons. For example, in the csv handler fetch step, it means that there is either something in the CSV data that the pipeline can’t process (usually UTF-8 characters) or the file type isn’t a plain CSV. In a Quickbase search step it could mean a data set that is too large or special characters are causing an issue. Usually, Google Cloud Platform (GCP) has more information on these errors that will point you in the right direction.

Text displayed in this error: Something went wrong executing a step. Try modifying your pipeline.

Quickbase errors

These errors tend to be the easiest to troubleshoot and test. The example below is a data type mismatch. This error is from trying to input a string into a numeric field.

Text displayed in this error: Validation error: Invalid literal for Decimal: 'Search Problem'

This error is from trying to input a non-unique value into a field required to be unique in the field properties.

Text displayed in this error: Remote service reported an error: 51 : Trying to add a non-unique a value to a field marked unique : The field named "Custom Numeric Key" with field id 9 requires a 9 unique value.

These are just a few of the many Quickbase channel errors, but you will likely get the hang of these errors faster than all the other types above.

Pipeline call depth exceeded

You see the Pipeline call depth exceeded warning in your log when your pipeline is using the Quickbase or Callable Pipelines channel and has called or triggered itself more than 100 times. You should examine your pipeline and check to see if it has entered a loop.

Unable to show all options

In this example, tables aren't displaying and this error displays:

Text displayed in this error: We're unable to show all options at this time. If you don't see what you're looking for, reload this page. If you need additional support, please contact us.

In this situation, try reauthenticating your pipeline’s user token access. See Manage User Tokens for more informatiion.