Is there a driver that supports Hive? The connection using SparkSQL directly cannot obtain metadata.
Apache has a bunch of Hive-compatible SQL frontends.
Try Trino or Druid.
ai told me we still need this package, right? But I didn't find the resources
metabase-sparksql-deps-1.2.1.spark2-standalone.jar
Metabase ships with everything needed for Spark SQL.
Don’t trust AI, it’s a liar.
No longer want to add components to the cluster ,The problem has been temporarily resolved by starting the Spark Thrift Server and connecting with SparkSQL. However, there is an issue with partitioned table parsing, where partition fields are repeatedly recognized as two identical fields. One has to manually insert the partition fields into the metabase_field table, but it's so cumbersome. For every partitioned table added in the future, one has to manually insert the partition fields first.