SUMMARY

Adds required OS dependencies for PyHive/Impyla(Impala) to work with user/password authentication.

Only adds 500KB to final container size.

docker inspect -f "{{ .Size }}" superset:latest
1450281124
docker inspect -f "{{ .Size }}" superset:fix-sasl-module
1450709446

Motivation: Allows connecting to Ascend.io using Superset over TLS with username and password.
impala://<service_account_username>:<service_account_password>@<ascendEnvironment>.sql.ascend.io:10000/<dataService>?auth_mechanism=PLAIN&use_ssl=true

TEST PLAN

1. Build existing superset master:
   sudo docker build -t superset:latest https://github.com/apache/superset.git#master
2. Build this PR branch
   sudo docker build -t superset:fix-sasl-module https://github.com/danielewood/superset.git#fix-sasl-module
3. With current superset:master, attempt to use authentication_method=PLAIN with Impala or auth=CUSTOM with PyHive
   sudo docker run --rm -it -u 0 --entrypoint bash superset:latest
4. See error from both PyHive and Impyla:
   thrift.transport.TTransport.TTransportException: Could not start SASL: b'Error in sasl_client_start (-4) SASL(-4): no mechanism available: No worthy mechs found'
5. With this PR patch, attempt to use authentication_method=PLAIN with Impala or auth=CUSTOM with PyHive
   sudo docker run --rm -it -u 0 --entrypoint bash superset:fix-sasl-module
6. See authentication error if you dont have a hive cluster to test against and use ascend with invalid credentials, otherwise see no errors because the connection worked.

Example:

sudo docker build -t superset:fix-sasl-module https://github.com/danielewood/superset.git#fix-sasl-module
sudo docker run --rm -it -u 0 --entrypoint bash superset:fix-sasl-module

# If you dont have a hive instance to test against, install socat to tunnel TLS to Ascend.io
# replace hive2.example.local with localhost
# apt update \
#    && apt install -y socat \
#    && socat TCP-LISTEN:10000,fork,reuseaddr openssl:trial.sql.ascend.io:10000,verify=1 &

python3 - <<-"EOF"
from sqlalchemy import create_engine
create_engine("hive://username:password@hive2.example.local:10000/database?auth=CUSTOM",echo=True,echo_pool='debug').connect()
EOF

####

# If you dont have a hive instance to test against, 
#  replace hive2.example.local with trial.sql.ascend.io, 
#  and add ";use_ssl=true" to the end of the connection string

pip install impyla
python3 - <<-"EOF"
from sqlalchemy import create_engine
create_engine("impala://username:password@hive2.example.local:10000/database?auth_mechanism=PLAIN",echo=True,echo_pool='debug').connect()
EOF

ADDITIONAL INFORMATION

• Brings desired results from abandonded PR FIX Hive Database connection - TTransportException: TSocket read 0 bytes #7015
• Fixes Hive connection  #2174
• Fixes SQLAlchemy URI for hive problem #2632