Thanks for your rely.
`SELECT * FROM tb WHERE id = 1` can be converted into `UNION ALL`, but i am worried how to handle some aggregation SQLs, e.g `SELECT AVG(NUM) FROM tb`. Juan Pan panj...@apache.org Juan Pan(Trista), Apache ShardingSphere On 10/22/2019 11:04，Julian Hyde<jh...@apache.org> wrote： Not currently, but it wouldn’t be too much work. Consider a similar query: SELECT * FROM mysqlTable UNION ALL SELECT * FROM cassandraTable This would convert into an EnumerableUnion which would send sub-queries to the two back ends and combine the results. You’d need a new relational operator which, I assume, would go with whichever result arrives first. A new sub-class of RelNode, perhaps similar to EnumerableUnion, or perhaps you could use a table-valued function. Julian On Oct 21, 2019, at 7:27 PM, Juan Pan <panj...@apache.org> wrote: Hi everyone, Thanks for your attention. I can not get a clear result after read most of Calcite document. So i send this email for your suggestion. Suppose there are two data storages, e.g, MySQL and Cassandra behind Calcite, and data is separately stored in two of them, can i execute a query, e.g `SELECT * FROM tb WHERE id = 1` simultaneously on two of data storages through Calcite? In other words, i want to get the final combined result from MySQL and Cassandra, which store part of data in different forms separately through Calcite. Looking forward to your suggestions and thoughts. Best wishes, Trista Juan Pan panj...@apache.org Juan Pan(Trista), Apache ShardingSphere
