I'm facing an error while running following code. Please help me solving this issue or suggest an alternative way to connect Vertica server via JDBC.
I have also tried connecting using pyspark. But with pyspark we are not able to execute sql query in vertica server.
System Configuration:
Linux with python 3.7.0
jaydebeapi version - 1.1.1
jpype version - 0.7.2
import jaydebeapi
import jpype
jdbc_url = 'jdbc:vertica://XX.XX.XXX.XXX/DATA?XXXXX'
vertica_properties={"user": "XXXX", "password": "XXXXX"}
jars = ['XXXX/vertica-jdbc-9.3.1-0.jar',
'XXXX/vertica-spark2.1_scala2.11.jar',
'XXXX/CustomVerticaDialect.jar']
connection = jaydebeapi.connect(jclassname='com.vertica.jdbc.Driver', url=jdbc_url, driver_args=vertica_properties, jars=jars)
connection
ERROR
--------------------------------------------------------------------------
AttributeError Traceback (most recent call last)
<ipython-input-5-cc7d2752ecd8> in <module>()
----> 1 connection = jaydebeapi.connect(jclassname='com.vertica.jdbc.Driver', url=jdbc_url, driver_args=vertica_properties, jars=jars)
2 connection
/BDAPP/python/anaconda3/lib/python3.7/site-packages/jaydebeapi/__init__.py in connect(jclassname, url, driver_args, jars, libs)
379 else:
380 libs = []
--> 381 jconn = _jdbc_connect(jclassname, url, driver_args, jars, libs)
382 return Connection(jconn, _converters)
383
/BDAPP/python/anaconda3/lib/python3.7/site-packages/jaydebeapi/__init__.py in _jdbc_connect_jpype(jclassname, url, driver_args, jars, libs)
180 types = jpype.java.sql.Types
181 types_map = {}
--> 182 for i in types.__javaclass__.getClassFields():
183 types_map[i.getName()] = i.getStaticAttribute()
184 _init_types(types_map)
AttributeError: type object 'java.sql.Types' has no attribute '__javaclass__'
The jaydebeapi seems to be incompatible with jpype 0.7.2 as of Mar 2020. Here is the github page that points out this issue: http://github.com/baztian/jaydebeapi/issues/99.
The solution as per the github page is to downgrade jpype to jpype 0.6.3 using either:
pip install JPype1==0.6.3 --force-reinstall
or
conda install -c conda-forge JPype1=0.6.3 --force-reinstall
Downgrading jpype worked for me