Hi everyone, Did anyone ever encounter the following error while trying to create a catalog with iceberg table sink using a Glue catalog implementation?

org.apache.flink.table.api.ValidationException: Could not find any factory for identifier 'iceberg' that implements 'org.apache.flink.table.factories.CatalogFactory' in the classpath

I am trying to connect to an iceberg table hosted on AWS Glue catalog from a Flink application running on AWS Managed Service for Flink. The SQL statement I am running:

CREATE CATALOG glue_catalog WITH (
        'type'='iceberg',
        'warehouse'='<s3://my-bucket/prefix>',
        'catalog-impl'='org.apache.iceberg.aws.glue.GlueCatalog',
        'io-impl'='org.apache.iceberg.aws.s3.S3FileIO'
);

I bundled the two necessary jar files (

iceberg-flink-runtime

and

iceberg-aws-bundle

as per the iceberg-aws integration docs) into an uber jar and checked that the

iceberg

identifier exists in that uber jar:

Available factory identifiers are:

blackhole
datagen
filesystem
kinesis
iceberg
print

But when I try to run the flink app it doesn’t seem to find the

iceberg

identifier’s implementations and I get the above-mentioned error. Does anyone know how can I resolve this?

First check release notes and make sure that Iceburg & AWS Glue versions are compatible with your flink version. Read release notes carefully.

Ok it turns out I was a bit stupid. I was trying to run my application code locally first to see if I would have any issues before deploying it to Managed Flink, but I guess the local flink python driver is different from the aws one because when I ran it on AWS Managed Flink I did not get this error.

`yeah there can be some differences when moving onto AWS