How does the driver communicate with HBase?
The driver uses the Stargate REST API to communicate with HBase. You must have Stargate enabled and running in HBase before the driver can successfully connect.
Why do some operations, such as aggregations and joins, take longer when using HBase directly than when HBase is used through Phoenix?
HBase does not expose rich enough APIs for the driver to be able to have the data source perform the necessary functionality, so it must pull the data back and perform the operations itself. Phoenix adds coprocessors to HBase to allow SQL processing within the cluster and thus can return only the result of the operation which in many cases is much smaller than the raw data that is used to create the result.
Why do some columns return incorrect data through the driver?
HBase does not store any metadata about the data within it, it simply stores the bytes. The driver attempts to detect the original type of the data but may be incorrect depending on the format of the data. Simply use the schema editor to correct the types for the affected columns and save the schema to return correct data.
Why does Tableau take so long to connect and switch to a worksheet?
Tableau must probe each ODBC driver to determine what capabilities it has so it can tailor the SQL and behavior it uses with that driver. The way it does this can cause excessive work to be done and cause extended periods of inactivity. To avoid this, install the TDC file that comes with the driver to pre-configure Tableau for use with the driver.
Does the driver support write-back?
No, the driver is read-only
Does the driver work with Apache Knox
Yes, the driver will work with Apache Knox