Jump to: navigation, search

Installing SpeechMiner

This section describes how to prepare your environment before you install SpeechMiner, and how to install the SpeechMiner software. The SpeechMiner components can be installed on one or more servers, as required by the particular configuration of your system.

Getting Started

Before you install SpeechMiner, make sure of the following:

Then, review the Pre-installation Checklist before you begin the installation process.

System Requirements

For more information about the supported operating environments, see the Supported Operating Environment Reference Guide.


Each server in the system should have at least the following amounts of available disk space before installing SpeechMiner:

  • All servers: Approximately 1 GB of disk space for the recognition engine
  • Recognition server(s): For the UPlatform service, 20 MB of disk space for the runtime folder plus approximately 10 GB for caching recognition packages. (The exact amount required for caching depends on the size of the implementation.)
  • Database server: At least 20 GB for the SpeechMiner database, plus 20 MB of disk space for the runtime folder of the UConnector service. In addition, on some types of recording-system integrations that have a very high volume, a larger (10 GB-200 GB) storage area is needed for temporary files.
Important
  • The initial size of the database is about 20 GB; it may grow larger, depending on the call volume and the call-purging policy.
  • On relatively high-volume installations, UConnector may need its own dedicated server.
  • Web server: About 20 MB for the SpeechMiner virtual folder, plus additional space for cached reports and call audio. (The exact amount required for caching depends on user activity.)
  • Interaction Receiver: About 15 MB for the Interaction Receiver virtual folder.
  • Machines running SMART: About 1 GB of disk space for the recognition engine
  • The database must run on a machine on which one of the following SQL servers is installed:

    • Microsoft SQL Server 2008 with Reporting Services, SP1 or above (Enterprise edition is recommended for large installations.)
    • Microsoft SQL Server 2012 with Reporting Services

    All machines must have Windows operating systems. The following table lists the Windows operating systems that are compatible with each SpeechMiner component.

    Component Compatible OS
    Servers (platform, database, and Web)
    • Windows Server 2008 R2 64 bit
    • Windows Server 2012
    SMART
    • Windows Server 2008 R2 64 bit
    • Windows Server 2012
    • Windows 7 Professional
    • Windows 8 Professional
    Web interface
    • Windows XP Professional (SP3 or above)
    • Windows Vista Business 32 bit
    • Windows 7 Professional 32 bit
    • Windows Server 2008 R2

    The SpeechMiner web interface is compatible with Internet Explorer (IE) versions 8, 9, and 10, and Google Chrome.

    Neither JAVA, .NET, or ActiveX are required for web-based access to SpeechMiner (via the SpeechMiner web-based interface).

    Users of the SpeechMiner web application must have a functioning audio device on their desktop. Users browsing using Internet Explorer should have Windows Media Player version 10 or 11 installed.

    Machines running SpeechMiner servers and applications should have at least the following amounts of memory:

    • Database server: 4 GB - 32 GB
    • Web server: 4 GB
    • SMART application: 1 GB minimum; 2 GB recommended
    • Platform server: 4 GB - 8 GB (allow 1 GB per recognizer task, as a rule of thumb)

    What Is Installed?

    The following software will be installed on the machines in your system:

    Machine Components
    On All Servers
    On the Recognition Server(s)
    On the DB Server
    • MS-SQL 2008 or 2012 Server, including MS Reporting Services (normally installed by the customer beforehand)
    • SpeechMiner database
    • UConnector service
    On the Web Server
    • SpeechMiner virtual folder
    On the Interaction Receiver Server
    • Interaction Receiver virtual folder
    On Every Machine Running SMART

    Ports Used by the System Components

    The following ports are used by SpeechMiner:

    Important
    The ports listed are the default ports. Most of them can be changed upon request.
    Source Destination Protocol and Port Pair (ex. TCP 3389)
    Web servers, Platform servers, SpeechMiner Administrator Workstations (SMConfig/SMART) Database server tcp 1433
    Database server, Web server, Platform servers, SpeechMiner Administrator Workstations (SMConfig/SMART) MS-SQL report server http 80 / https 443
    SpeechMiner Administrator Workstations (SMConfig/SMART) Web servers http 80 / https 443
    Genesys Interaction Recording server Interaction Receiver Server http 80 / https 443
    Web Servers Web servers http 80 / https 443
    SpeechMiner Administrator Workstations (SMConfig) Platform servers, Web servers tcp 135
    Platform servers (recognition), SpeechMiner Administrator Workstations (SMART) SpeechMiner Nuance License server tcp 27000 + another port (can be configured in license file)
    Web servers, Platform servers, SpeechMiner Administrator Workstations (SMConfig/SMART) File System smtp over tcp 445
    Web servers, SpeechMiner Administrator Workstations (SMConfig/SMART) Active Directory tcp 88
    Web servers, Platform servers Email server smtp over tcp 25

    Ports and Protocols Required for SpeechMiner UConnector

    SpeechMiner UConnector requires access to the recording-system database and file-storage system. Access to the database is implemented using the TCP protocol with port 1433. (The port number can be configured on the database server.) Access to the file-storage system is implemented using SMB over TCP protocol with port 445. Other protocols can be used as well, if they are available in the underlying file-storage system.

    This page was last edited on May 4, 2014, at 10:16.
    blog comments powered by Disqus