topic Re: Does Qlik Replicate read the transaction log when SQL Server MS-CDC is the source? in Qlik Replicate

Does Qlik Replicate read the transaction log when SQL Server MS-CDC is the source?

MoeE — Wed, 15 Apr 2026 23:59:49 GMT

Hi,

When a Qlik Replicate task has SQL Server as the source with MS-CDC, is Qlik reading the transaction log or is it reading the change tables created by MS-CDC?

I know when MS-Replication is enabled, QR will read the transaction log. Is it different with MS-CDC.

Regards,

Mohammed

Re: Does Qlik Replicate read the transaction log when SQL Server MS-CDC is the source?

MoeE — Thu, 16 Apr 2026 00:05:39 GMT

Hi,

It looks like I've found the answer. According to the SQL Server (MS-CDC) source endpoint documentation

"When replicating data from Microsoft SQL Server using MS-CDC, changes are captured from Microsoft Change Tables as opposed to the SQL Server transaction log"

Thanks,

Mohammed

Re: Does Qlik Replicate read the transaction log when SQL Server MS-CDC is the source?

john_wang — Thu, 16 Apr 2026 00:23:35 GMT

Hello Mohammed, @MoeE

This is correct. Thanks for the update.

Best Regards,

John.

Re: Does Qlik Replicate read the transaction log when SQL Server MS-CDC is the source?

MoeE — Thu, 16 Apr 2026 00:30:02 GMT

Hi John,

Can you please confirm something for me? If SQL Server using MS-CDC is the source, can QR capture and apply updates to tables without any primary keys or indexes on the source table?

The documentation seems to suggest this in the MS-Replication SQL Server source documentation

Thanks,

Mohammed

Re: Does Qlik Replicate read the transaction log when SQL Server MS-CDC is the source?

john_wang — Thu, 16 Apr 2026 01:08:42 GMT

Hello Mohammed, @MoeE ,

You’re right — a primary key or unique index is not mandatory for MS-CDC. However, performance can degrade significantly, especially when the target table is large. This is because the target database must locate the correct row for updates, and without a primary key or unique index, this process becomes much less efficient. Defining a primary key or unique index helps speed up row identification and improves overall performance.

Regards,

John.