IPython/Jupyter notebooks for Apache Impala

with Kerberos authentication

1. Connect to the target database

### requires Cloudera impyla package and thrift_sasl
### edit the value kerberos_service_name as relevant for your environment



In [1]:

    
from impala.dbapi import connect
conn = connect(host='impalasrv-prod', port=21050, kerberos_service_name='impala', auth_mechanism='GSSAPI')

2. Run a query and fetch the results



In [2]:

    
cur = conn.cursor()

cur.execute('select * from test2.emp limit 2')

cur.fetchall()









    Out[2]:





[(7876,
  'ADAMS',
  'CLERK',
  7788,
  datetime.datetime(1987, 5, 23, 0, 0),
  1100,
  None,
  20),
 (7499,
  'ALLEN',
  'SALESMAN',
  7698,
  datetime.datetime(1981, 2, 20, 0, 0),
  1600,
  300,
  30)]

Integration with pandas



In [3]:

    
cur = conn.cursor()

cur.execute('select * from test2.emp')

from impala.util import as_pandas
df = as_pandas(cur)

df.head()

More examples of integration with IPython ecosystem



In [4]:

    
cur = conn.cursor()
cur.execute('select ename, sal from test2.emp')
df = as_pandas(cur)

%matplotlib inline
import matplotlib
matplotlib.style.use('ggplot')

df.plot()









    Out[4]:





<matplotlib.axes._subplots.AxesSubplot at 0x7fc66cfd8ed0>



In [ ]:

	empno	ename	job	mgr	hiredate	sal	comm	deptno
0	7698	BLAKE	MANAGER	7839.0	1981-05-01	2850	NaN	30
1	7900	JAMES	CLERK	7698.0	1981-12-03	950	NaN	30
2	7566	JONES	MANAGER	7839.0	1981-04-02	2975	NaN	20
3	7782	CLARK	MANAGER	7839.0	1981-06-09	2450	NaN	10
4	7902	FORD	ANALYST	7566.0	1981-12-03	3000	NaN	20