QR: Replicating to AWS S3 resulting with error Windows/WinHTTP error code is 12002
Hi team,
We are shifting out Qlik Replicate servers from an OnPrem data centre to GCVE.
Or replication tasks that are writing to AWS S3 started returning an intermittent error message:
00031576: 2026-09-01T12:41:32 [FILE_FACTORY ]E: Send request failed: Windows/WinHTTP error code is 12002: The operation timed out . [1000100] (AWS_SDK_CPP Internal Logger:337) 00031576: 2026-09-01T12:41:32 [FILE_FACTORY ]E: Encountered network error when sending http request. [1001705] (anw_retry_strategy.cpp:170)
This error message happens from all our servers writing to S3 and it has an intermittent nature. For example I can click "test connection" and it will fail the first two times and then work the third. It almost seems like the connection needs to be warmed up.
Our network team speculates the timeout is resulting from a DNS request is not responding in time.
We have tried updating the cdcIdleTimeout on the endpoint to a very high number; but this hasn't solved the problem.
Since the S3 connector is written on the AWS SDK - is there a way to potentially access properties like httpSocketTimeout through the endpoint to give Qlik more time waiting for a response?
Or any other idead if people have come across this problem before?
Hi, There could be an internal parameter in the S3 endpoint for this. Recommend you engage with Qlik Support for a faster answer. However, I feel this best solved in the Windows/GCVE/AWS networking stack rather than working around it in Replicate. Dropped TCP connections is a common reason so check the KeepAlive parameters. Cheers!
