Data connectors are a powerful new feature introduced with JMP 18. Given recent updates to JMP’s support for SAS and SharePoint, these connectors are often essential. But data connectors are not always required. This article explores each data connector type to help you determine when they are a "must-have" versus an optional tool.
SAS / SAS Viya (Required)
Connecting to SAS in JMP now requires a data connector. While advanced users can use the JMP Python environment as an alternative, most users will find it significantly easier to configure a SAS Local, SAS Remote, or SAS Viya data connector.
Snowflake (Recommended)
While Snowflake connections traditionally use Snowflake DSII ODBC drivers, the driver often presents reliability issues or requires expensive third-party licenses (like the CData driver). JMP now provides a Snowflake data connector that utilizes REST API technology to connect. This method is free, significantly faster, and more reliable than any ODBC driver.
SharePoint & OneDrive (Optional/Recommended)
Historically, users accessed SharePoint via the JSL Open() function. However, as JMP evolved, this method became prone to failure.
- The connector: The OneDrive & SharePoint data connector offers a stable, dedicated connection.
- The alternative: If your computer is already mapped to OneDrive at the OS level, you can still use File > Open. However, the data connector is the preferred method.
The OneDrive & SharePoint data connector is available at the JMP Marketplace.
ODBC (Optional/Context-Dependent)
ODBC was the primary driver behind the development of data connectors. While standard ODBC drivers cover basic parameters (server, port, password), many database management systems (DBMS) require specialized configurations to run efficiently.
- When to use: Use an ODBC data connector if you need to manage specific behaviors like SQL Server’s Block Fetch, DB2’s Scrollable Cursors, or Databricks Quote Types. These allow you to fine-tune performance and avoid query errors.
- When to skip: If you can currently run queries via Query Builder or Open Table without errors, a data connector is unnecessary. There is no functional difference in how JMP handles standard ODBC traffic if these specialized parameters aren't required.
- Plus one: If you do need a connector, it is easier to build it from an existing DSN configured on your machine rather than starting from scratch.
You must be a registered user to add a comment. If you've already registered, sign in. Otherwise, register and sign in.