Share via

Internal Server Error Liquid Transformation Logic App Consumption Big Files

José Gomes 20 Reputation points
2026-02-18T15:06:11.9533333+00:00

Hello,

Since February 1st, we have been experiencing issues with a Consumption-based Azure Logic App that uses a Liquid template to transform a large JSON message.

Previously, the transformation would complete successfully within a few minutes. However, it is now consistently failing after approximately 20 minutes with an "Internal Server Error".

Key details:

  • The Logic App workflow has not changed.

The Liquid template has not been modified.

The input files are the same as those that were successfully processed before.

The failure occurs specifically during the Liquid transformation step.

The runtime duration before failure is around 20 minutes.

Given that nothing in the workflow or payload structure has changed, we would like to understand:

Has anything changed recently in the backend of the Integration Account service?

  1. Are there updated limits regarding Maximum message size for Liquid transformations?

We would appreciate your assistance in identifying whether this is related to a platform change, a newly enforced limit, or a service degradation.

Thank you in advance for your support.

Azure Logic Apps
Azure Logic Apps

An Azure service that automates the access and use of data across clouds without writing code.

{count} votes

Answer accepted by question author
  1. Praveen Kumar Gudipudi 1,880 Reputation points Microsoft External Staff Moderator
    2026-02-19T13:01:12.3833333+00:00

    Hello José Gomes,

    Consumption workflows run on shared infrastructure managed by Microsoft.

    Over time:

    Regional usage increases

    Workloads are redistributed

    Available CPU/memory per execution varies

    If your Liquid transform was already near the execution ceiling, even a small reduction in available compute can turn:

    3–4 minutes → 20 minutes → timeout

    No workflow change is required for this to happen.

    Please accept as answer and do a Thumbs-up to upvote this response if you are satisfied with the community help. Your upvote will be beneficial for the community users facing similar issues.

    0 comments No comments

2 additional answers

Sort by: Most helpful
  1. José Gomes 20 Reputation points
    2026-02-19T11:11:07.4133333+00:00

    Hi Praveen

    The Region is West Europe.
    The message is around 40MB.
    This is the template (I changed the fields names)

    [
    {% for item in collectionA %}
        {
            "fieldA1": "{{item.fieldA1}}",
            "fieldA2": "{{item.fieldA2}}",
            "fieldA3": "{{item.fieldA3}}",
            "fieldA4": "{{item.fieldA4}}",
            "fieldA5": "{{item.fieldA5}}",
            "arrayA1":[
                {% for subItemA in item.arrayA1 %}
                {
                    "fieldB1": "{{subItemA.fieldB1}}",
                    "fieldB2": "{{subItemA.fieldB2}}",
                    "fieldB3": "{{subItemA.fieldB3}}"
                },
                {% endfor %}
            ],
            "fieldA6": "{{item.fieldA6}}",
            "fieldA7": "{{item.fieldA7}}",
            "fieldA8": "{{item.fieldA8}}",        
            "arrayA2":[
                {% for subItemB in item.arrayA2 %}
                {
                    "fieldC1": "{{subItemB.fieldC1}}",
                    "fieldC2": "{{subItemB.fieldC2}}",
                    "fieldC3": "{{subItemB.fieldC3}}",
                    "fieldC4": "{{subItemB.fieldC4}}"
                },
                {% endfor %}
            ],
            "arrayA3":[
                {% for subItemC in item.arrayA3 %}
                {
                    "fieldD1": "{{subItemC.fieldD1}}",
                    "fieldD2": "{{subItemC.fieldD2}}",
                    "fieldD3": "{{subItemC.fieldD3}}"
                },
                {% endfor %}
            ],
            "fieldA9": "{{item.fieldA9}}",
            "objectA1":
                {
                    "fieldE1": "{{item.objectA1.fieldE1}}",
                    "fieldE2": "{{item.objectA1.fieldE2}}"
                },
            "fieldA10": "{{item.fieldA10}}",
            "fieldA11": "{{item.fieldA11}}",
            "fieldA12": "{{item.fieldA12}}",
            "fieldA13": "{{item.fieldA13}}",
            "fieldA14": "{{item.fieldA14}}"
        },
    {% endfor %}
    ]
    

  2. Q&A Assist
    2026-02-18T15:06:25.3066667+00:00
    AI generated content. This question contains an answer created with AI from Q&A Assist. Learn more

    It appears that you are experiencing an Internal Server Error during the Liquid transformation step in your Azure Logic App, particularly when processing large JSON messages. While the workflow and Liquid template have not changed, there are a few considerations to keep in mind:

    1. Backend Changes: It is possible that there have been updates or changes in the backend of the Integration Account service that could affect performance or limits. However, specific details about recent changes are not provided in the context.
    2. Message Size Limits: While the context does not explicitly mention updated limits regarding the maximum message size for Liquid transformations, it is important to note that Azure Logic Apps may have inherent limits that could impact processing large files. If the size of your JSON message is at or near any existing limits, this could lead to failures.
    3. Service Degradation: If the service is experiencing degradation or increased load, this could also contribute to longer processing times and failures. The context suggests that performance can be affected by backend system load, especially in Consumption workflows.

    Given that the failure occurs consistently after approximately 20 minutes, it may be beneficial to monitor the Azure service status for any ongoing issues or to reach out to Azure support for more specific insights related to your Integration Account and Liquid transformation limits.


    References:

    0 comments No comments

Your answer

Answers can be marked as 'Accepted' by the question author and 'Recommended' by moderators, which helps users know the answer solved the author's problem.