Inconsistent table replication behavior in qlik replicate task (Few tables erroring out)

[sandeep_r]

Hi Qlik Team,

We are running into a weird issue where few of the tables are erroring out in a replication task. To be precise, the task is trying to load (full load later CDC) ~150 tables and around 2 to 3 tables are erroring with the below exception. 
We have checked with our db team and these are not partitioned tables.
When we reloaded the tables using the reload option on the ui, qlik was able to load them.
Question, Why is qlik running into this odd behavior. It is concerning as we have no clue of why this is happening. Manual intervention is not an option given the number of task that are being planned.
We also have a concern with CDC after the full load. Is it possible that this issue might arise on tables after full loads when the tables are in CDC.

stack trace:
00099692: 2025-05-14T14:18:24 [SOURCE_CAPTURE ]W: Table 'dbo.<TABLE>' is not uniformly mapped across partitions. Therefore - it is excluded from CDC (sqlserver_log_metadata.c:1733)
00099692: 2025-05-14T14:18:24 [SOURCE_CAPTURE ]I: Table 'dbo.<TABLE>' has been mapped and registered for CDC. (sqlserver_log_metadata.c:1104)
00100632: 2025-05-14T14:18:24 [TASK_MANAGER ]W: Table 'dbo'.'REC_PeriodInformation' (subtask 0 thread 0) is suspended. (replicationtask.c:3239)

[Dana_Baldwin]

Hi @sandeep_r 

1. Please advise the full version number of Qlik Replicate you are using.

2. What is your source (and target) endpoint and version? If SQL Server, there are some queries we can run against the source to determine if indexes need to be rebuilt. In any event, please check for index fragmentation on the source and resolve that to rule it out.

3. What is the source driver/client version?

4. As the component showing this warning is Source_Capture it an occur during full load or CDC but seems most likely to happen during full load.

5. Are you using the Parallel Load feature for these tables?

6. Was any maintenance actions happening on the source at the same time? Best practice for this is to stop the Replicate task during maintenance activities (index defrag/rebiuld, etc.).

Thanks,

Dana

[Dana_Baldwin]

@sandeep_r Also, if this is SQL Server source:

https://help.qlik.com/en-US/replicate/November2022/Content/Global_Common/Content/SharedReplicateHDD/...

CDC is not supported from tables which have both compressed and uncompressed partitions.
If only some of the partitions are compressed, the following message will be shown: Table 'xxx' is not uniformly mapped across partitions. Therefore, it is excluded from CDC.

This may not be your situation since you could reload the tables, but trying to share everything I found on that warning message.

[john_wang]

Hello @sandeep_r ,

It's unexpected to encounter this error on a non-partitioned table. Could you please run the following SQL to confirm whether the table is actually partitioned?

SELECT SCHEMA_NAME(t.schema_id) AS SchemaName, *
FROM sys.tables AS t
JOIN sys.indexes AS i
ON t.[object_id] = i.[object_id]
JOIN sys.partition_schemes ps
ON i.data_space_id = ps.data_space_id
WHERE t.name = '<TableName>';

Replace <TableName> by your table name.

Regards,

John.

[sandeep_r]

Hi @john_wang , I have run the query you have given (with my table_name), The response was an empty dataset.

[srikanth-appampally]

Hi @Dana_Baldwin,

on Point 3,

• Microsoft SQL Server 2019
• ODBC driver version: '17.07.0002'

Target: GCS -> Snowflake

On Point 4,  it is happening during full load.

On Point 5, Yes we are doing parallel load. 

[Dana_Baldwin]

Thanks @srikanth-appampally 

Please advise on points 1 and 2 so we can confirm the version of the endpoints and drivers is supported by the version of Replicate you are using.

Thanks,

Dana