# Converting SQL_VARIANT on PDW


There are a number of data types that are currently unsupported on PDW. Among those are SQL\_VARIANT. The interesting part of that is that SQL\_VARIANT along with several other unsupported data types are found all over the place in system tables in PDW. There is a simple solution, convert the value to something like VARCHAR. Well this wouldn't be a blog post if that was all I had to say on the matter. Turns out that you can select the data just fine, convert it in a select just fine, but for some odd reason not do an insert with it just fine. Let's take the partitions range table for this example. `SELECT CONVERT(VARCHAR(200), value, 101) AS ConvertedText FROM sys.partition_range_values` Try to insert that into a temp table and bad things happen. `--Create a temp table CREATE TABLE #Partitions ( Value VARCHAR(200) ) WITH (LOCATION = USER_DB)` `--Load the temp table INSERT INTO #Partitions SELECT CONVERT(VARCHAR(200), value, 101) AS ConvertedText FROM sys.partition_range_values` `--Select the records SELECT * FROM #Partitions` `--Drop the table DROP TABLE #Partitions` You will get this error when it runs the insert statement even though it is clearly being converted. \[error\]Msg 100077, Level 16, State 1, Line 1 Unsupported data type error. Statement references a data type that is unsupported in Parallel Data Warehouse, or there is an expression that yields an unsupported data type. Modify the statement and re-execute it.\[/error\] The workaround: CTAS works like a charm! `CREATE TABLE TestPartition WITH (DISTRIBUTION = REPLICATE) AS SELECT CONVERT(VARCHAR(200), value, 101) AS ConvertedText FROM sys.partition_range_values` `SELECT * FROM TestPartition` `DROP TABLE TestPartition` For those who require visual proof: ![](https://cdn.hashnode.com/res/hashnode/image/upload/v1716523560799/7937c782-9e50-4dee-b27c-c5e7b0234a73.png) Happy CTASing and happy Friday!