Skip to main content

OLake Go Compatibility with Query Engines

You can query OLake Go Iceberg tables from multiple engines. The table below shows catalog compatibility at a glance, with links to setup guides.

Query ToolAWS GlueHive MetastoreJDBCREST GenericLakekeeperUnityPolarisS3 TablesNessieBigLakeDocs
Amazon Athena✅❌❌❌❌❌❌✅❌❌AWS Docs
Apache Spark (v3.3+)✅✅✅✅✅✅✅✅✅✅Spark Docs
Apache Flink (v1.18+)✅✅✅✅✅✅✅✅✅✅Flink Docs
Trino (v475 +)✅✅✅✅✅✅✅✅✅✅Trino Docs
Starburst Enterprise✅✅✅✅✅✅✅✅✅✅Starburst Docs
Presto (v0.288 +)✅✅✅✅✅✅✅❌✅❌Presto Guide
Apache Hive (v4.2.0)✅✅❌✅✅✅✅❌✅✅Hive Docs
Apache Impala (v5.0)❌✅❌✅✅❌✅❌✅❌Impala Docs
Dremio (v25/26)✅✅❌✅✅✅✅✅✅✅Dremio Docs
DuckDB (v1.2.1)✅❌❌✅✅✅✅✅✅✅DuckDB Docs
ClickHouse (v25.8 +)✅✅❌✅✅✅✅❌✅✅ClickHouse Docs
StarRocks (v4.0 +)✅✅✅✅✅✅✅✅✅✅StarRocks Docs
Apache Doris (v3.1 +)✅✅❌✅✅✅✅✅✅✅Doris Docs
BigQuery (BigLake)✅❌❌❌❌✅❌❌❌✅BigQuery Docs
Snowflake (GA)✅❌❌✅✅✅✅✅✅✅Snowflake Docs
Databricks (Spark)✅✅✅✅✅✅✅✅✅✅Databricks Docs
Databricks✅✅❌❌❌✅❌❌❌❌Databricks Docs
AWS Redshift✅❌❌❌❌✅✅✅❌❌Redshift Docs
Azure Synapse Analytics✅✅✅✅✅❌✅✅✅❌Synapse Docs
Symbol Guide
  • ✅ — Query Engine connects to this catalog and reliably reads what OLake Go writes
  • ❌ — Query engine doesn't support this catalog type

Important Notes​

  • Presto → REST catalog supported from v0.288+; Glue requires the AWS SDK fat-jar.
  • Apache Hive v4.2.0 → Glue requires adding the AWS bundle.
  • Impala → Reliable support for Hive Metastore; limited REST catalog support (read only).
  • ClickHouse → Full read support; experimental Iceberg write support introduced in v25.7+
  • Snowflake → Can read external REST catalogs, but tables are read-only for external catalog workflows.
  • Azure Synapse Analytics → Iceberg support via Spark pools requires manual runtime configuration.

Delete Format Support by Query Engine​

OLake Go writes Iceberg tables using one of three row-level delete formats: equality deletes, positional deletes, or deletion vectors. Not all query engines support all three formats, so check the table below to see which query engines support which delete file format.

Query EngineEquality DeletesPositional DeletesDeletion Vectors
Amazon Athena✅✅❌
Apache Spark (v3.3+)✅✅✅
Apache Flink (v1.18+)✅✅✅
Trino (v475+)✅✅✅
Starburst Enterprise✅✅✅
Presto (v0.288+)✅✅❌
Apache Hive (v4.2.0)✅✅❌
Apache Impala (v5.0)✅✅❌
Dremio (v25/26)⚠️✅⚠️
DuckDB (v1.2.1)✅✅✅
ClickHouse (v25.8+)✅✅✅
StarRocks (v4.0+)✅✅❌
Apache Doris (v3.1+)✅✅✅
BigQuery (BigLake)✅✅✅
Snowflake (GA)❌✅✅
Databricks (Spark)✅✅✅
Databricks❌❌✅
AWS Redshift✅✅❌
Azure Synapse Analytics✅✅✅
Dremio delete format caveats
  • Equality deletes: Dremio does not support reading global equality deletes on Iceberg v2 tables when the delete file’s partition spec is unpartitioned.
  • Deletion vectors: Supported on Dremio Cloud only; Dremio self-managed (self-hosted) does not support reading deletion vectors.
Delete Format Notes
  • Equality deletes and positional deletes are Iceberg v2 row-level delete formats.
  • Deletion vectors is Iceberg v3 row-level delete format.