Pipeline Failure in ADF

Veizi Ioanna 10 Reputation points
2023-08-01T08:07:52.2166667+00:00

Hi,

I have created & scheduled some pipelines in ADF and they sometimes fail and sometimes run successfully.

These pipelines include activities that I copy data from oracle db to Azure SQL db.

The error I get is that we loose connection with Azure SQL db.

Could you please advice why is this happening ?

Thanks a lot.

Azure Data Factory
Azure Data Factory
An Azure service for ingesting, preparing, and transforming data at scale.
{count} votes

2 answers

Sort by: Most helpful
  1. Konstantinos Passadis 19,686 Reputation points MVP
    2023-08-01T08:40:19.56+00:00

    Hey Ioanna !

    Here is a similar thread

    https://learn.microsoft.com/en-us/answers/questions/707576/how-do-i-pause-an-azure-sql-database-(serverless)

    Hope this helps !

    In case there is not Auto Pause we need to investigate further the issue !

    Kindly mark any answer that helped as Accepted and Upvote !

    0 comments No comments

  2. QuantumCache 20,676 Reputation points Moderator
    2023-08-02T17:29:38.25+00:00

    Hello @Veizi Ioanna

    Thanks for sharing the more info on the error! That was helpful to have an initial diagnosis!

    The error message indicates that the pre-login handshake failed, which could be caused by a network issue or a firewall issue.
    Make sure that the firewall settings for the Azure SQL Database allow the integration runtime to access the database. You can check the firewall settings in the Azure portal.

    Did you try to involve anyone from your SQL Server DBA team? in your project?

    Please do let us know if you need any help.

    Dt:08/30/2023: Below is the Resolution Shared by Veizi Ioanna

    I edited the activities in the pipelines to 'retry' in case they fail, when the disconnections occur and the 2nd time they are executed they complete successfully. For now this is the solution I have come up to.

    Please click "Accept Answer" . So that we can close this case.


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.