This topic describes the hardware and software configurations supported in TIBCO StreamBase® and TIBCO® Live Datamart release 10.3.0.
TIBCO StreamBase supports Windows and macOS platforms for development, and supports Windows and Linux platforms for production deployment. MacOS platforms are not supported for build or production servers. In all cases, these are 64-bit platforms running on Intel or AMD x86_64 hardware.
The StreamBase Studio development environment is supported on:
The StreamBase Runtime for build or staging servers and for production deployment is supported on the following platforms:
On Red Hat platforms, TIBCO supports and recommends Red Hat Enterprise MRG Realtime for deployments that require low-latency processing.
The StreamBase Runtime is explicitly not supported for build, staging, or production deployment on:
In all cases, TIBCO StreamBase supports the latest service packs and operating system updates.
**Users of macOS 10.12 Sierra and High Sierra 10.13, see Starting StreamBase Studio on macOS Sierra.
|Processor Architectures||StreamBase runs on Intel and AMD 64-bit x86_64 processors. For all Intel processors, hyperthreading is supported. For Intel and AMD processors, multiple CPU cores are supported and encouraged.|
|JDK requirements for running applications||
StreamBase Studio and the StreamBase Runtime require an installation of Oracle JDK 8 for Studio and for running Maven at the command line.
On all supported operating systems, the StreamBase installation kit provides the runtime components of Oracle JDK version 1.8.0_171 in a subdirectory of the StreamBase installation directory, in order to run StreamBase program fragments and applications from within StreamBase Studio.
The mvn command shipped with StreamBase runs with Java and must be able to locate a java command in the PATH or by means of JAVA_HOME. Using a StreamBase Command Prompt on Windows or a StreamBase configured shell on macOS or Linux allows the Oracle JDK shipped with Studio to be used for command line operations as well.
If you have a site policy that requires the use of a particular JDK version, see the installation requirements for an external JDK in Installing a Command Line JDK.
Follow these guidelines to select a JDK for system-level installation:
|JDK Cryptography Features||Due to licensing restrictions, the version of the Oracle JDK bundled with TIBCO StreamBase does not contain Elliptical Curve
Cryptography functionality. This is a feature of Oracle Java, available from a Java Cryptography Architecture provider.
In particular, if your site’s Kerberos server infrastructure uses stronger encryption levels (such as AES-256) than the ones supported by the standard Oracle JDK shipped with StreamBase, you may need to download and install the Java Cryptography Extension Unlimited Strength package from Oracle. Consult with your site's Kerberos administrators for site-specific requirements.
|JDK requirements for compiling client applications||To compile your own StreamBase Java clients , custom functions, or custom operators with the StreamBase Java Client library, your development host must have Oracle JDK 8. TIBCO recommends that you compile with the same JDK version as the one bundled with your StreamBase installation, described in the previous rows.|
|LiveView Web version||TIBCO LiveView Web Standard Edition is included with TIBCO Live Datamart. This provides the runtime component of LiveView
*For production servers, determine the appropriate RAM amount from peak-load and volume testing on your staging servers.
|Screen resolution||To use StreamBase Studio, your monitor's screen resolution should be at least 1280 x 768 pixels. Higher resolution settings and multiple monitor configurations are strongly recommended.|
|Web browsers||Web browsers supported for viewing the online StreamBase documentation, are:
There is an exception for the .NET Client API documentation, which is only provided on Windows systems. To view this documentation without complications, you must use Firefox or a Microsoft-provided browser. This page of the documentation describes the workarounds available for Webkit-based browsers such as Chrome or Opera.
On all platforms, expect the installed TIBCO StreamBase directory to require one to two GB.
The installer files require around one GB each, and may require several hundred more MB for temporary files during installation.
Disk use for your application workspace depends on the volume of data that you store. Expect to need another one or two GB.
Running a StreamBase EventFlow fragment in the StreamBase Runtime creates a node directory that consumes an amount of disk space equivalent to the amount of RAM specified for that node to use. Running fragments from the command line creates one node directory per node, which are temporary and can be removed after a fragment successfully runs.
|Perl||Perl is required by OpenSSL on Linux and macOS platforms. Perl is shipped with macOS, but barebones installations of Linux may not have it. Be sure to install Perl on all Linux StreamBase hosts. Perl is not required on Windows platforms.|
|Zip||On RHEL, CentOS, and Oracle Linux platforms, make sure the zip command is installed and available in the PATH.|
|Linux compatibility library||On newer Linux distributions, the Linux package to support GCC 3.3 compatibility must be installed to support a Thomson Reuters
library included with TIBCO StreamBase:
|External JDBC-compliant databases||TIBCO supports interaction with JDBC-compliant (2.0 or later) external databases:
StreamBase applications can interact with supported external databases as follows:
|Native database connectivity||TIBCO StreamBase supports interaction with certain databases using a high-performance interface that uses the database vendor's
connectivity API instead of JDBC. The event-store database for which StreamBase provides native connectivity support is:
|Building StreamBase C++ custom clients on Linux and macOS||TIBCO StreamBase requires GCC 4.1.x through GCC 4.5 for building C++ clients on Linux. For further information, see Creating C++ Clients in the API Guide.
StreamBase does not support building clients with GCC 4.6 or newer, which are the default compilers on newer Linux distributions. On such distributions,
install GCC and G++ 4.5 from your distribution's repository, and set the
To build C++ clients on macOS, install the latest Xcode version and the Command Line Tools for Xcode released for your macOS version. (These are free downloads, but require an Apple ID.) The Command Line Tools package provides
|Building StreamBase C++ custom clients on Windows||
To compile C++ StreamBase client applications for Windows, TIBCO StreamBase supports:
See Configuring Visual C++ for instructions on configuring your Visual C++ projects for use with the StreamBase C++ API.
|Building .NET clients on Windows||To develop StreamBase client applications in C# .NET, TIBCO StreamBase supports Visual Studio 2010 (deprecated), 2012, or 2013 for C# development.|
|.NET framework version requirements||To run .NET client applications, the target machine must have the .NET framework 4.0 or later, plus a pair of StreamBase assemblies and supporting DLLs that are installed as part of your StreamBase installation. For target machines that do not have StreamBase installed, you only need to distribute the two assemblies and two Microsoft DLLs along with your .NET client application to run, as described in Deploying .NET Clients.|
|Python Client API||The StreamBase Python Client API supports Python 2.6 or 2.7 on Linux and Windows only. For Windows, install a third-party Python distribution such as ActiveState Python. Make sure the python command is on the PATH.|
|Python Support for Python Operators||The TIBCO StreamBase® Operators for Python, and Python Instance are both compatible with the Python versions described in Python Compatibility.|
Adapters are programs that convert data to and from the StreamBase tuple protocol.
Embedded adapters execute in the same StreamBase Runtime process that runs your StreamBase program fragment. The base TIBCO StreamBase installation includes a large collection of embedded adapters, described in the Adapters Guide.
Most embedded adapters are Java-based, and run on all platforms that support the StreamBase Runtime. Embedded adapters that make calls to third-party native libraries are restricted to the platforms for which such libraries are provided.
The TIBCO StreamBase® Add-in for Microsoft Excel lets you use a StreamBase or Live Datamart application in combination with an Excel spreadsheet to process tuple data in real time and to display the data dynamically. The Excel Add-in is bidirectional and can both receive data from an application and publish data to a running application.
The Add-in is only supported on Windows platforms. To use it, you must have Excel 2007, 2010, 2013, or 2016.
The Add-in is distributed in both 32-bit and 64-bit versions; match these to the bitness of your Microsoft Office installation, not the bitness of your Windows installation.
The Excel Add-in is packaged separately from TIBCO StreamBase. If you have a license for it, you can download, install, and run the Add-in. To acquire a license, contact your TIBCO Sales representative.
|TIBCO Rendezvous messages||StreamBase supports TIBCO Rendezvous with a set of embedded adapters.
The adapters take advantage of the
|TIBCO ActiveSpaces||The StreamBase operators for TIBCO ActiveSpaces support ActiveSpaces 2 releases, starting with version 2.1.|
|TERR||TIBCO® Enterprise Runtime for R Developer Edition included with StreamBase is TERR release 4.4. The Developer Edition is licensed for developer evaluation purposes only.|
Thomson Reuters RMDS
|StreamBase adapters for Thomson Reuters support RMDS 5 and RMDS 6 infrastructure components. Consult with your TIBCO Software
representative for the latest information on Thomson Reuters API support.
Because of limitations in third-party libraries, StreamBase running on 64-bit Linux or 64-bit Windows does not support SASS3 connections to Thomson Reuters RTIC servers.
|Apache Hadoop Libraries||The Apache Hadoop libraries used by several StreamBase adapters is version 8.2.1.|
|Apache Kafka components||The Apache Kafka adapter suite is built against version 0.11.0.0 of the Apache Kafka libraries.|
|Spark MLlib library version||Tested primarily with Spark MLlib version 2.|
|H2O library version||Tested with H2O library version 184.108.40.206|
|JPMML library version||Tested with JPMML library versions 1.3.7, PMML specification versions 3.0, 3.1, 3.2, 4.0, 4.1, 4.2 and 4.3|