Skip to main content

SQLAlchemy support for Splice Machine RDBMS

Project description

Splice Machine SQLAlchemy Adapter

Description:

  • This python package allows users to integrate new and existing SQLAlchemy applications with Splice Machine
  • SQLAlchemy exposes a uniform API for database-backed applications, which can be easily used with Splice Machine via this adaptor
  • Supported on Splice Machine Versions >= 2.8.0.1920
  • Requires Python 3.6+

Installation:

There are two ways to install this Driver. With and without ODBC Driver custom installation. We recommend without. When running these commands, you must either use sudo or have root access.

Without ODBC Driver Custom Installation

Mac

brew install unixodbc pip install splicemachinesa

Linux

First ensure you have the following packages installed. They are necessary for pyodbc.
Ubuntu:

  • build-essential
  • libssl-dev
  • libffi-dev
  • unixodbc
  • unixodbc-dev
  • python3-dev
  • curl
  • unzip
  • iodbc

sudo apt -y update && sudo apt -y upgrade && sudo apt -y install build-essential libssl-dev libffi-dev unixodbc unixodbc-dev python3-dev curl unzip iodbc

Centos:

  • unzip
  • gcc
  • openssl-devel
  • gcc-c++
  • unixODBC-devel
  • unixODBC
  • python3-devel

sudo yum -y update && sudo yum -y install unzip gcc openssl-devel gcc-c++ unixODBC-devel unixODBC python3-devel

Then run pip install splicemachinesa

Note: If you have an error during installation, it is likely with the installation of PyODBC. When searching the error, reference PyODBC in the search.

With Custom ODBC Driver Installation

ODBC Driver:

First, download the appropriate Splice Machine ODBC driver for your system.



Mac OSX 64 bit Driver: Download
Linux 64 bit Driver: Download

Note: Windows is not currently supported for Splice Machine SQLAlchemy ODBC


Then, follow the instructions here to finish the installation, but use the download links above to retrieve the Driver tarball

Once the Driver is installed, you can pip install splicemachinesa.

NOTE: If you are on Mac, you must install unixODBC
brew install unixodbc

pip install splicemachinesa

Usage

You can use this package for SqlAlchemy usage or raw ODBC usage

ODBC Connection Only

from splicemachinesa.pyodbc import splice_connect
ODBC_CONNECTION = splice_connect(UID=[UID], PWD=[PWD], URL=[URL], SSL=[SSL])

Filling in UID, PWD, URL, and SSL with the proper values for your database. SSL defaults to 'basic' If you are connecting to the Splice Database inside the same network (ie standalone splice) you will set SSL=None

SqlAlchemy

There are three URL formats that can be used to access SpliceMachine via ODBC through SQLAlchemy. The first will suffice in most cases.


1: Standard URL format

Format: splicemachinesa://[user]:[password]@[host]:[port]/[database]

The Driver Name defaults to SpliceODBCDriver, which is the default name when the driver is installed (specified in odbc.ini). However, this name can be overridden through the SPLICE_ODBC_DRIVER_NAME environment variable.

Example:

from sqlalchemy import create_engine
url = 'splicemachinesa://splice:admin@127.0.0.1:1527/splicedb'
engine = create_engine(url)

2. Custom Configuration Without Driver

Format: splicemachinesa://?DRIVER=[driver]&URL=[URL]&PORT=[PORT]&UID=[USER]&PWD=[PASSWORD]

You can use this SQLAlchemy driver without an ODBC configuration (meaning running the installation from Splice Machine docs) by manually specifying the driver path.

  • On OSX (given installation with sudo ), the Driver is found in /Library/ODBC/SpliceMachine/libsplice_odbc64.dylib
  • On OSX, Without sudo this driver is located at $HOME/Library/ODBC/SpliceMachine/libsplice_odbc64.dylib
  • On Linux (given ODBC Driver installation with sudo), this is found in /usr/local/splice/libsplice_odbc.so
  • On Linux, Without sudo it is found in $HOME/splice.

We provide a utility function for simplifying the URL building process.
This method is easier for an automated driver configuration (e.g. inside Docker) because you only need to copy to Driver binary.

Note: Error messages will not be rendered properly with this approach.

Example Linux:

from sqlalchemy import create_engine
from splicemachinesa.utilities import url_builder

url = url_builder('/usr/local/splice/lib/libsplice_odbc.so', host='localhost', port=1527,
user='splice', password='admin')
 
engine = create_engine(url) 

Example Mac:

from sqlalchemy import create_engine
from splicemachinesa.utilities import url_builder

url = url_builder('/Library/ODBC/SpliceMachine/libsplice_odbc64.dylib', host='localhost', port=1527, 
user='splice', password='admin')
 
engine = create_engine(url) 

3. DSN Configuration

Format: splicemachinesa://[dsn] Splice Machine SQLAlchemy also supports ODBC DSNs for Driver configuration. This means that rather than explicitly specifying configuration options in the URL string, they are rendered from the odbc.ini file. This is also the only method that supports Kerberos Authentication for Splice Clusters. You can see how to use Kerberos here. The odbc.ini file is located at these locations depending on whether or not the installer was run as root.

With/without sudo installation- Mac OSX: $HOME/.odbc.ini
With sudo Linux: /etc/odbc.ini
Without sudo Linux: $HOME/.odbc.ini. In this file, if you have a key named USER, rename this to UID.

The default DSN is SpliceODBC64.

Example:

from sqlalchemy import create_engine
url = 'splicemachinesa://SpliceODBC64'
engine = create_engine(url)

Testing

  1. First make sure you have a fresh installation of Splice Machine running (either Standalone or Cloud/Bespoke/On-Prem) with an appropriate SQLAlchemy Splice Machine connection string for accessing it
# register any changes for testing
cd /path/to/splice_sqlalchemy
sudo pip install .
sudo pip install pytest
py.test -vv --db-url 'splicemachinesa://[...]'

Features not yet supported

  • Common Table Expressions
  • Indices Reflection
  • Check + Unicode Reflection
  • Nullable Reflection
  • ORDER BY COLLATE
  • Empty Set Insertion
  • Unicode String Support
  • Reflector for Order by Foreign Key
  • limit_offset_in_unions_from_alias

Project details


Release history Release notifications | RSS feed

This version

0.3.3

Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Files for splicemachinesa, version 0.3.3
Filename, size File type Python version Upload date Hashes
Filename, size splicemachinesa-0.3.3.tar.gz (30.5 kB) File type Source Python version None Upload date Hashes View

Supported by

Pingdom Pingdom Monitoring Google Google Object Storage and Download Analytics Sentry Sentry Error logging AWS AWS Cloud computing DataDog DataDog Monitoring Fastly Fastly CDN DigiCert DigiCert EV certificate StatusPage StatusPage Status page