Epicor ERP Integration - Queries on how to setup MS Azure SQL Database (SaaS) as source via MS-CDC

[harikesh_1991]

Hello @john_wang @DesmondWOO @Dana_Baldwin ,
Hope you are doing good.

As part of a requirement, I need to setup a source endpoint which is off the type MS Azure SQL Database (SaaS) to do CDC(Insert/Updates/Deletes) and replicate it to a target which is an Oracle endpoint.
Since the source endpoint is a SaaS version of MS Azure SQL Database instance, MS Replication is not an option to setup. However, MS-CDC can help to do CDC and transfer the data.

However, could you please advise what are the setup that needs to be requested to the source team to setup at their end so that we can facilitate MS-CDC for it.

Could you please advise if we need to request that the Qlik user being used for connecting to their source database to have full privilege or just a read/write and DDL access enough for it?

Also could you please advise if they need to enable MS-CDC on the Azure SQL database hosting Epicor?
EXEC sys.sp_cdc_enable_db;

Additionally, do they have to enable CDC for all the required tables to be replicated?

EXEC sys.sp_cdc_enable_table
@source_schema = 'dbo',
@source_name = 'TableName',
@role_name = NULL,
@supports_net_changes = 1;

Additionally, I believe we need to get the below details from the source team to setup the endpoint connection. Could you please confirm on the same and kindly let me know if I am missing out any detail to be requested?

Server:
Username : This should be our qlik username right?
Password: Password for Qlik user.
Database Name:

Currently for all the other CDC between Oracle to Oracle, we have used a standby DR server dedicated for CDC, and hence it helped in reducing the latency drastically. Will a MS Azure SQL server instance(SaaS) have a standby DR server that we can utilize as a source endpoint to ensure we use it solely for CDC purpose ? Or is it possible to request them to create one and use it solely for CDC purpose?

I was under the assumption that since we use the MS-CDC option, we would be reading the changes directly from the change tables created at the source and not from the transaction logs. Please correct me if I am wrong, so in this case, can we use a standby DR server for log shipping between primary live DB to standby DR DB and then read changes from the change tables there?

It would be really helpful if I can get any info on this.

Thanks in Advance,

Regards,
Harikesh OP

[DesmondWOO · Accepted solution]

Hi @harikesh_1991 ,

1. Qlik User Privileges
Please refer to the following link:

https://help.qlik.com/en-US/replicate/November2025/Content/Global_Common/Content/SharedReplicateHDD/...

2. Enable CDC
Yes, but please note that our guide does not explicitly mention "@supports_net_changes = 1;". Enabling this parameter will impact the audit trail of updates. Please check the statement below:

https://help.qlik.com/en-US/replicate/November2025/Content/Replicate/Main/MicrosoftAzureSQLMS-CDC/az...

3. Endpoint connection
It depends on your authentication method settings. Please check

https://help.qlik.com/en-US/replicate/November2025/Content/Global_Common/Content/SharedReplicateHDD/...

4. DR Database
From my understanding, SQL Server's CDC (Change Tables) are stored and managed within the Primary database, which is architecturally different from Oracle's standby mining. Since this is a native Microsoft feature, I recommend consulting a Microsoft Engineer for the possibility.

Regards,
Desmond

[harikesh_1991]

Thanks a lot for your insights on this. @DesmondWOO . 🙂