This chapter explains how to configure servers and the services to run on them within VoiceObjects Desktop.
Configuring a Server
To configure a new server, do the following:
In the Object Palette, open the Configuration folder and double-click Server. The Server editor will open up in the editor area.
In Desktop for Web you open up the Server editor by selecting Configure New Server from the Tools menu.
Note that creating new Server objects requires a controller or administrator role.
Enter a name for the new Server object in the Name field at the top of the editor window.
The Configuration Parameters section contains the following fields:
· Reference ID
Specifies the reference ID of the server.
Note that in order for the installation to work properly, the reference ID specified here needs to be identical with the <servername> specified in the VOServer_Configuration.xml file. The default value entered by the installer is VOServer, which corresponds to the default value for a new Server object. For more information see Server name in Chapter 1 Advanced Configuration of VoiceObjects in the Administration Guide.
This parameter is mandatory.
8Caution: Only alphanumeric characters (a-z, A-Z, 0-9) and underscores are allowed. No special characters like blanks, ?, !, %, <, > etc., and no characters coming from character sets other than ISO-8859-1 can be used. Note that the ID is not case-sensitive, so you cannot have two reference IDs in your project that only differ in case.
8Caution: Never change this value for a server that has active server instances. In particular, never change this value in an operational deployment environment. Service outages may occur as a result.
· Session guarantee
Specifies the guaranteed number of concurrent dialog sessions available to this server. Guaranteed sessions are set aside in a separate pool that can only be accessed by the respective server. This ensures that the session partitioning defined for a specific server can be met. For more information see Chapter 9 Service Level Agreements (SLA).
Note that if the number of guaranteed sessions specified here exceeds the number of available sessions licensed for the VoiceObjects installation, it is limited accordingly.
This parameter is optional. If it is not defined, no guaranteed number of concurrent sessions is set aside for this server.
To activate changes made to the session guarantee, the server needs to be reset.
· Session limit
Specifies the maximum number of concurrent dialog sessions available on this server. If this number is reached, then additional attempts to create a dialog session for any service hosted on this server are rejected. In this case, the caller hears or sees a configurable message indicating that the service is currently not available. Once the number of concurrently active sessions drops below the limit, new calls are accepted again. For more information see Chapter 9 Service Level Agreements (SLA).
This parameter is optional. If it is not defined, the maximum number of concurrent sessions for this server coincides with the number of concurrent sessions licensed for the VoiceObjects installation.
To activate changes made to the session limit, the server needs to be reset.
· Control Center URL
Specifies the URL to be used to access the Control Center for this server. Typically, it is of the form http://server:port/VoiceObjects/Services/WSProvider?wsdl.
To simplify configuration, a default value for localhost is provided.
In a cluster configuration, the URL may point to any available server instance in the cluster.8Caution: This value is crucial to be able to open the Control Center for a server.
· Certificate URL
Specifies the path to a SSL certificate, which is used for authentication if the Control Center URL defined above uses an HTTPS connection. The path can either be defined by using the HTTP or file protocol.
· Testing URL
Specifies the URL to be used to initiate Debug Viewer or Phone Simulator sessions for this server. Typically, it is of the form http://server:port/VoiceObjects/.
To simplify configuration, a default value for localhost is provided.
For a change to take effect, the Control Center must be closed and re-opened.
· Enable System DB logging
Indicates whether System DB logging is enabled for this server. If it is disabled, individual service settings are overridden and dialog statistics are not logged to the Infostore Repository. In addition, attempts by Log objects to write to the destination System DB are ignored.
This setting can be overridden from within the Control Center.
By default, this check box is selected.
Note that System DB logging is only available if Infostore is licensed.
· Enable Custom DB logging
Indicates whether Custom DB logging is enabled for this server. If it is disabled, individual service settings are overridden and attempts by Log objects to write to the destination Custom DB are ignored.
This setting can be overridden from within the Control Center.
By default, this check box is selected.
Note that Custom DB logging is only available if Infostore is licensed.
· Enable utterance recording
Indicates whether utterance recording is enabled for this server. If it is disabled, individual service settings are overridden and no utterance recording files are created.
In the Hosted Services section the set of services that are hosted on this server is listed. See Configuring a Service for more details.
@8 Tip: In a server farm scenario, it is necessary to use IP-based load balancing in between the media platform and the server farm. Depending on the specific needs, either hardware-based or software-based load balancers can be used. Refer to the documentation of the respective products for more information on their configuration. More information can also be found in Appendix A Sizing and Configuration Recommendations in the Administration Guide.
i8 Note: If a Server object has been locked by some other user, neither its settings (including its reference ID) nor the list of hosted services can be modified. The Server object first needs to be unlocked.
For more information on the locking and unlocking of objects, refer to Chapter 7 Basic Commands in the Desktop for Eclipse Guide or Desktop for Web Guide.
Configuring a Service
To configure a new service, do the following:
In the Object Palette, open the Configuration folder and double-click Service. The Service editor will open up in the editor area.
In Desktop for Web you open up the Service editor by selecting Configure New Service from the Tools menu.
Note that creating new Service objects requires a controller or administrator role.
Enter a name for the new Service object in the Name field at the top of the editor window.
There are four sections to specify application parameters, communication parameters, logging parameters, and default parameters.
To activate changes made to these settings, the service needs to be redeployed. See Service Manager in Chapter 3 Managing Servers and Services for details.
Application parameters
Application parameters specify the basic behavior of the service including the start object of the application, the language, and the input mode.
The Application Parameters section contains the following fields:
· Start object
Specifies the start object of the application associated with this service. Typically, this is a Module object. When the service is requested by the media platform at call time, the start object is processed, initiating the dialog flow enclosed within it.
For XDK applications, the start object may be defined using a URL pointing to a start object within an XDK file, e.g. http://myserver/PrimeInsurance.xml#rootModule. The reference ID of the object must be used to identify it, not the name.
In Desktop for Web, the Context Menu button 
to the right of the Start object field is only available when working within a project.
This parameter is mandatory.
· Language
Specifies the initial language the service runs in. This parameter can be overwritten at call time by passing the parameter vsLanguage in the initial URL request from the media platform that initiates the dialog. See Service URL Configuration in Chapter 4 Service Deployment for details.
The default setting Default indicates that the default language specified for the driver is used. If no default language is specified for the driver the default language specified for the server is used. For more information on the server default language see the Installation Guide.
Appendix A Language Codes in the Object Reference contains a list of all language codes available in VoiceObjects together with the respective language they represent.
· Input mode
Specifies the initial active input mode for the dialog. This parameter can be overwritten at call time by passing the parameter vsInputMode in the initial URL request from the media platform that initiates the dialog. See Service URL Configuration in Chapter 4 Service Deployment for details.
Voice indicates that caller input can only be made by voice.
DTMF indicates that caller input can only be made by DTMF.
Voice+DTMF indicates that caller input can be made by either voice or DTMF.
The default setting is Voice+DTMF.
8Caution: When changing the input mode to either Voice or DTMF you need to make sure that you specify the required grammars and outputs for the respective input modes.
When working in text-based channels, the input mode is automatically set to Voice through a switch in the corresponding driver definition.
· Enable validation
Indicates whether the server should validate the integrity of the application when loading the service either upon start-up or when the service is redeployed. It is strongly recommended to enable validation in order to detect application problems. This also enables the automatic safeguard capabilities, which leave a deployed application intact instead of replacing it with a broken one.
By default, this check box is selected.
8Caution: If this check box is clear, applications may be deployed that fail at call time due to structural problems.
· Retain caches on shutdown
Indicates whether the active and restore caches for this service should be retained even when the server is completely shut down. If selected, the application definition is not newly loaded when the server starts up the next time, but instead restored from the persisted cache in the Metadata Repository. This option is particularly relevant for volatile services that can be created using the Web Services Interface. For more details, refer to the Web Services Guide.
By default, this check box is clear.
i8Note: When changing this setting on a service that is already deployed on a server, it is necessary to redeploy the service to activate the change. This is true both when activating the retain option (to create the retained caches) as well as when deactivating it (to remove the retained caches).
Communication parameters
Communication parameters provide the essential information to connect the media platform to the server including the service name, and the media platform driver to be used.
Depending on a users role, this section may be inaccessible.
The Communication Parameters section contains the following fields:
· VSN (reference ID)
VoiceObjects Service Name. Specifies the unique name of the service. The VSN can be used to identify a service when initiating it by means of an HTTP request from the media platform. See Service URL Configuration in Chapter 4 Service Deployment for details.
8Caution: Only alphanumeric characters (a-z, A-Z, 0-9) and underscores are allowed. No special characters like blanks, ?, !, %, <, > etc., and no characters coming from character sets other than ISO-8859-1 can be used. Note that the ID is not case-sensitive, so you cannot have two reference IDs in your project that only differ in case.
8Caution: Note that the VSN must not end in either _shadow or _inactive, as the server uses these strings internally.
This parameter is mandatory.
i8Note: For information on how to change the VSN of a deployed service see Maintenance Operations in Chapter 3 Managing Servers and Services.
· DNIS mapping
Specifies a comma-separated list of phone numbers that are mapped to this service. See Service URL Configuration in Chapter 4 Service Deployment for details.
This parameter is optional.
· Startup mode
Specifies what is to be done with this service when a server hosting it is started.
Automatic indicates that the service is started together with the server.
Manual indicates that the service is not started together with the server, and shows up in the list of hosted services as stopped.
Disabled indicates that the service is not started together with the server, and shows up in the list of hosted services as disabled.
The default setting is Automatic.
· Cache standby
Specifies whether the cache for this service may dynamically be dropped from memory when there is insufficient space to hold all service caches in memory simultaneously. For details on cache configuration refer to Cache Management in Chapter 1 Advanced Configuration of VoiceObjects in the Administration Guide.
On demand indicates that the cache may be dropped.
Never indicates that the cache must always be kept in memory.
The default setting is On demand.
i8Note: It is recommended to use Never very sparingly since VoiceObjects Server is typically better positioned to decide which services to keep in memory.
· Driver
Specifies the media platform that the service runs on. This parameter can be overwritten at call time by passing the parameter vsDriver in the initial URL request from the media platform that initiates the dialog. See Service URL Configuration in Chapter 4 Service Deployment for details.
The default setting Default indicates that the default driver selected for the server is used. For more information on specifying the default driver see the Installation Guide.
· Respect user-agent in Web channel
If the check box Respect user-agent in Web channel is selected the server reads out the user-agent header information in the initial HTTP request coming from the client; this is a string that identifies the browser so that Web servers can respond with optimized code. Based on an up-to-date list of Web browser user-agents the server tries to map the agent to one of the available VoiceObjects Web drivers. If the mapping succeeds, then this driver is taken for the session instead of the driver defined in the Service object or in the service URL parameter vsDriver.
Since the Web channel is the only channel for which the requesting platform (=browser) is unknown before the call comes in, this setting is required to be able to dynamically select a suitable Web driver that presents the application in an optimal way.
If no corresponding driver can be found in the lookup phase or the driver is not from the Web channel, the driver setting from the service or URL parameter is taken and the user-agent is ignored.
This parameter can be overwritten at call time by passing the parameter vsRespectUserAgent in the initial URL request from the media platform that initiates the dialog. See Service URL Configuration in Chapter 4 Service Deployment for details. If both vsRespectUserAgent and vsDriver are provided in the URL the server first tries to do a mapping of agent to driver, and if that fails it takes the vsDriver value.
By default, this check box is clear.
· Session guarantee
Specifies the guaranteed number of concurrent dialog sessions available to this service. Guaranteed sessions are set aside in a separate pool that can only be accessed by the respective service. This ensures that the session partitioning defined for a specific service can be met.
Note that if the number of guaranteed sessions specified here exceeds the number of guaranteed sessions for the site this service belongs to or for the server it is hosted on, the number of guaranteed sessions is limited accordingly. For more information see Chapter 9 Service Level Agreements (SLA).
This parameter is optional. If it is not defined, no guaranteed number of concurrent sessions is set aside for this service.
· Session limit
Specifies the maximum number of concurrent dialog sessions allowed for this service. If this number is reached, then additional attempts to create a dialog session for this service are rejected. In this case, the caller hears a configurable message indicating that the service is currently not available. Once the number of concurrently active sessions drops below the limit, new calls are accepted again. For more information see Chapter 9 Service Level Agreements (SLA).
This parameter is optional. If it is not defined, the maximum number of concurrent sessions for the service coincides with the maximum number for the server it is hosted on.
· Session distribution
Specifies the distribution of concurrent dialog sessions among services within a site and on a given server.
The session distribution specifies a percentage indicating the allocation of available dialog sessions to the respective service. Note that this setting is only relevant if a session limit is defined for the site.
The sum of all percentage values assigned to services within a single site must not exceed 100. If it does, all session distribution settings for this site are ignored. For more information see Chapter 9 Service Level Agreements (SLA).
The default setting is Disabled.
Instead of selecting one of the values from the drop-down list, it is also possible to type in a custom value. In this case, the value must lie between 0 and 1. E.g. to select a session distribution of 27%, the value 0.27 needs to be entered.
Logging parameters
Logging parameters specify the logging behavior of the service including the default logging destination, the list of activated logging categories, and the activation settings for database logging.
The Logging Parameters section contains the following fields:
· Enable System DB logging
Indicates whether System DB logging is enabled for this service. If it is disabled, dialog statistics are not logged to the Infostore Repository. In addition, attempts by Log objects to write to the destination System DB are ignored.
This setting can be overridden from within the Control Center.
By default, this check box is clear.
· EnableInputStatelogging
Indicates whether System DB logging writes detailed information for each individual input state visited by a caller. If it is disabled, only a single data record is written per call indicating information such as ANI, DNIS, length and time of call, etc. Also business task, layer and module related data is stored in the Infostore Repository.
For further information, refer to the Infostore Guide.
By default, this check box is clear.
· Enable Custom DB logging
Indicates whether Custom DB logging is enabled for this service. If it is disabled, attempts by Log objects to write to the destination Custom DB are ignored.
This setting can be overridden from within the Control Center.
By default, this check box is clear.
· Coverage
Specifies the percentage of dialogs that should be written to System DB logging (and, optionally, Custom DB logging; see Filter scope below).
In deployments with high call volumes, it is sometimes desirable to not write out detailed statistics on each and every call, but only on a fraction thereof. Coverage can be selected from 10% to 100% of all dialogs. Each individual dialog is then logged either completely or not at all. The coverage selection algorithm uses a uniform distribution on dialog sessions to minimize the impact of coverage filtering on subsequent reporting.
The default is 100%.
Instead of selecting one of the values from the drop-down list, it is also possible to type in a custom value. In this case, the value must lie between 0 and 1. E.g. to select a coverage of 27%, the value 0.27 needs to be entered.
· Filter scope
Specifies whether the coverage defined above extends only to input states, to System DB logging (the default), or to System DB and Custom DB logging.
If coverage is less than 100% and System DB and Custom DB is selected, then dialogs that do not write to System DB logging will also not write to Custom DB logging. This option should be chosen if Custom DB logging is used as an extension of System DB logging.
If coverage is less than 100% and System DB is selected, then dialogs that do not write to System DB logging will still write to Custom DB logging. This option should be chosen if Custom DB logging is used e.g. to trigger back-end processing (such as making an entry into an ordering system).
If coverage is less than 100% and InputStateis selected, then dialogs always write a single record but only write detailed Input State logging in the selected percentage of cases (but any individual dialog either writes all input states or none).
The default is System DB.
· Enable utterance recording
Indicates whether utterance recording is enabled for this service. If it is disabled, no utterance recording files are created.
By default, this check box is clear.
· Utterance recording path
Specifies the path that utterance recording files are written to.
For more information see Chapter 6 Recording of Utterances.
· Utterance recording URL
Specifies the URL through which utterance recording files can be retrieved.
For more information see Chapter 6 Recording of Utterances.
· Coverage
Specifies the percentage of dialogs that should create utterance recording files.
In deployments with high call volumes, it is sometimes desirable to not write files for each and every call, but only for a fraction thereof. Coverage can be selected from 10% to 100% of all dialogs. The coverage selection algorithm uses a uniform distribution on dialog sessions to minimize the impact of coverage filtering on subsequent reporting.
The default is 100%.
Instead of selecting one of the values from the drop-down list, it is also possible to type in a custom value. In this case, the value must lie between 0 and 1. E.g. to select a coverage of 3%, the value 0.03 needs to be entered.
For more information see Chapter 6 Recording of Utterances.
· Enable OSDM Event Logging
Indicates whether OSDM Event Logging is enabled for this service.
If enabled, the OSDM Event Log Destination field can be used to define a file to which the data is logged. If not defined, the default file as defined in the OSDM configuration is used. Note that paths in the OSDM Event Log Destination field need to use forward slashes, even on Windows.
By default, this check box is clear.
· Default logging destination
Specifies the default location for Log objects to write to.
Custom DB indicates that the information is written to a custom database table.
Log File indicates that the information is written to a server log file that can be accessed through the ControlCenter.
Log OSDM indicates that the information is written to the OSDM log.
Media Platform indicates that the information is written to the media platform log by using the <log> element in VoiceXML.
System DB indicates that the information is written to a special field within the System DB logging.
For more information, refer to the Log object in the Object Reference.
This parameter can be overwritten at call time by passing the parameter vsLoggingDestination in the initial URL request from the media platform that initiates the dialog. See Service URL Configuration in Chapter 4 Service Deployment for details.
The default setting is Log File.
· Logging categories
Specifies which log categories are active for this service. Multiple check boxes can be selected, and only Log items that match one of the selected categories will be activated at call time. This makes it easy e.g. to integrate debugging messages into an application for development and testing purposes, and then to disable those messages with a single click as the application goes into deployment.
Info is intended for informational messages that do not necessarily indicate error conditions, but provide information on the dialog processing.
Debug is intended for messages only used during application development and debugging. This category should typically be disabled for deployed services.
Error is intended for messages indicating error situations in the application. This category should typically be enabled for deployed services.
Statistics is intended for messages that provide statistical information on how callers use the application.
Billing is intended for messages that are used for billing purposes.
The set of selected categories can be overwritten at call time by passing the parameter vsLoggingCategories in the initial URL request from the media platform that initiates the dialog. See Service URL Configuration in Chapter 4 Service Deployment for details.
By default, all check boxes are selected.
· Maintainer e-mail
Specifies an e-mail address that error messages are sent to by the media platform.
This parameter is optional.
i8 Note: System and Custom DB logging are only available with a valid Infostore license.
Default parameters
Default parameters specify the defaults for a variety of service settings including e.g. barge-in and dialog exit type. Default parameters come in three groups:
· those that are relevant for the entire dialog and that can be changed by use of expression functions;
· those that are relevant for specific types of objects and that can be overridden locally within such objects;
· and those that are relevant for the entire dialog and that cannot be changed at call time.
The Default Parameters section contains the following fields:
· Grammar type
Specifies the standard grammar type. All Grammar items with type setting Default are controlled by this setting, so that the grammar type can be changed throughout an entire service with a single switch.
SRGS ABNF indicates W3C speech recognition grammar format in BNF notation.
Nuance GSL indicates Nuance GSL grammar format.
JSGF indicates Java speech grammar format.
SRGS XML indicates W3C speech recognition grammar format in XML notation.
Cisco Regular Expression indicates Cisco regular expression format.
Precompiled indicates precompiled grammar format.
None indicates that no standard grammar format is used.
Built-in indicates built-in grammars as defined in the VoiceXML specification.
The default setting Default indicates that the default grammar type for the selected driver is used.
Note that if a grammar type is selected which is not supported by the selected media platform, the platforms default type is used instead by the server.
· Video type
Specifies the standard video type. All Video objects with type setting Default are controlled by this setting, so that the video type can be changed throughout an entire service with a single switch.
Note that if a video type is selected which is not supported by the selected media platform, the platforms default type is used instead by the server.
· Barge-in
Specifies the standard setting for barge-in. All Output items with barge-in setting Default are controlled by this setting, so that barge-in can be enabled or disabled throughout an entire service with a single switch.
True indicates that barge-in is allowed.
False indicates that barge-in is not allowed.
The default setting is True.
· Grammar control
Specifies the standard setting for grammar control. All Input objects with grammar control setting Default are controlled by this setting, so that grammar control can be enabled or disabled throughout an entire service with a single switch.
True indicates that grammar control is enabled.
False indicates that grammar control is disabled.
The default setting is False.
For more information on grammar control, refer to Input in the Object Reference.
· Output mode
Specifies whether audio or TTS should be the preferred output mode.
The output mode is useful for testing whether all required audio files and alternative texts have been defined.
Audio:TTS indicates that audio is used if available, otherwise TTS.
TTS:Audio indicates that TTS is used if available, otherwise audio.
The default setting is Audio:TTS.
· Record utterances
Specifies whether utterances should be recorded for this service (True) or not (False). The default setting is True.
Note that this setting only has relevance if the setting Enable utterance recoding in the Logging Parameters section has been selected.
· Recording scope
Specifies the scope to be covered by utterance recording.
All indicates that all utterances are recorded.
No Match indicates that only utterances causing a No Match event are recorded.
Recognition indicates that only utterances leading to a successful recognition are recorded.
The default setting is All.
· Standby timeout
Specifies the time after which a dialog session is removed if there is no activity. As opposed to the dialog timeout, the standby timeout is applied during the processing of Plug-In, Transfer, or Recording objects. These are objects where some processing takes place outside of VoiceObjects, which can take longer than the usual dialog timeout. The standby timeout allows you to increase the timeout temporarily for these objects.
The default setting is 10 minutes.
Instead of selecting one of the values from the drop-down list, it is possible to type in a custom value. This value is interpreted as seconds. E.g. for a custom timeout of 5.5 minutes, type in 330.
· Dialog timeout
Specifies the time after which a dialog session is removed if there is no activity. Note that if you use call pausing (see Pause object in the Object Reference) or play long audio files to the caller, it might be necessary to increase this value in order to ensure that the dialog is not unintentionally terminated.
The default setting is 5 minutes.
For applications working in the text or Web channel, a shorter dialog timeout may be desirable, e.g. 40 seconds. In Input, Menu, Confirmation, and List objects, the dialog timeout can temporarily be increased in order to give callers more time to respond to them if required (e.g. when entering an address).
Instead of selecting one of the values from the drop-down list, it is also possible to type in a custom value. This value is interpreted as seconds. E.g. for a custom timeout of 1.5 minutes, type in 90.
· HTTP connector timeout
Specifies the time after which a HTTP request issued by a Connector object is aborted if no answer is received from the back-end.
The default setting is 10 Seconds.
Instead of selecting one of the values from the drop-down list, it is possible to type in a custom value. This value is interpreted as seconds. E.g. for a custom timeout of fifteen seconds, type in 15.
· Dialog exit type
Specifies what should be done when the end of the dialog is reached and the last object processed was not an Exit object (if it was, the exit type defined in that object is taken).
Disconnect indicates that the call is terminated with a hang-up by sending a <disconnect> tag to the media platform.
Exit indicates that control over the call is handed back to the media platform.
Return indicates that a <return> tag is sent to the media platform. This enables the continuation of an ongoing dialog that was initiated outside VoiceObjects.
The behavior in these cases is described by VoiceXML files that are sent to the media platform. The behavior can be modified by adjusting these files, which can be found under
..\VoiceObjects\Platform\Resources\System\VXML\<VoiceXML version>\
<Media Platform Provider>\
with the names Disconnect.vxml, Exit.vxml, and Return.vxml.
The default setting is Disconnect.
· Grammar precedence
Specifies the precedence in case both embedded and external grammar definitions are present in a Grammar object item.
Embedded:External indicates that an embedded definition takes precedence over an external grammar reference. External:Embedded indicates that the external grammar file will be taken rather than an embedded grammar definition.
The default setting is Embedded:External.
· Audio file extension
Specifies an audio file extension that overwrites all file extensions defined locally in Audio objects.
The default setting None indicates that the locally defined audio file extensions are used.
· Text file extension
Specifies a text file extension that overwrites all file extensions defined locally in Text objects.
The default setting None indicates that the locally defined text file extensions are used.
· Grammar file extension
Specifies a grammar file extension that overwrites all file extensions defined locally in Grammar objects.
The default setting None indicates that the locally defined grammar file extensions are used.
· Video file extension
Specifies a video file extension that overwrites all file extensions defined locally in Video objects.
The default setting None indicates that the locally defined video file extensions are used.
· Reject response
Defines the way in which the server responds to requests when no more licenses are available for the service.
Overflow indicates that a file overflow.vxml is sent, usually to indicate that the service is not available at the moment.
Redirect indicates that a file redirect.vxml is sent, usually to transfer the call e.g. to a call center agent.
The default setting is Overflow.
The behavior in both cases can be modified by adjusting the corresponding file, which can be found under
..\VoiceObjects\Platform\Resources\System\VXML\<VoiceXML version>\
<Media Platform Provider>\
The reject response behavior can be overwritten at call time by passing the parameter vsRejectResponse in the initial URL request from the media platform that initiates the dialog. See Service URL Configuration in Chapter 4 Service Deployment for details.
· Standard navigation
Standard navigation provides functionality such as Go back or Repeat. It is configured in the Module object by providing appropriate grammars for the various commands.
Disabled indicates that standard navigation is disabled, regardless of whether grammar entries are defined for it or not.
The default setting is to remember the last three dialog steps (3 Steps). This means that the caller can use the standard navigation command Back three times, and thus go back at most three steps in a dialog.
· HTTP proxy host
Specifies a proxy host to be used for all HTTP connectors (but not Web Service connectors) as well as external scripts referenced with the HTTP protocol. The default is not to use a proxy.
· HTTP proxy port
Specifies a proxy port to be used for all HTTP connectors (but not Web Service connectors) as well as external scripts referenced with the HTTP protocol. The default is not to use a proxy.
· Session variables
Specifies a comma-separated list of media platform session variables. This can be used for two purposes:
1 If a media platform calls the server through a custom VoiceXML page using <subdialog> and the platform wants to submit parameters to the application using <param> elements, define all required parameter names in this field.
2 If you want to fetch variables from the platform session context other than those already specified in the driver (like ANI, DNIS, etc.), define all additionally required variables with syntax vo_name (mp_name) in this field, where vo_name is the parameter name you want to use within the application and mp_name is the name used by the media platform.
Example: ANI (session.connection.remote.uri)
In both cases, the variables can be accessed through the expression function SESSION(vo_name).
· Resource path
Specifies a physical path that is used in Resource Locator objects using the option Allow service overwrite.
The default is not to use a resource path.
· Resource URL
Specifies a URL that is used in Resource Locator objects using the option Allow Service overwrite.
The default is not to use a resource URL.
· Configuration URL
Specifies the location of a configuration file that defines application default values. Both HTTP and file URLs are allowed. Note that it is required to mention the protocol (http:// or file://) in the URL.
See Application Defaults in Chapter 4 Service Deployment for details.
The default is not to use a configuration URL.
· OSDM URL
Specifies the URL that points to an installation of OpenSpeech DialogModules (OSDMs). This value overrides the one defined in the <osdmURL> element within the VOServer_Configuration.xml.
This parameter is only relevant if OSDMs have been licensed.
· Loop abort handler URL
Specifies the location of a file holding valid VoiceXML code that will be put in an event handler of type error for level 10 on each rendered VoiceXML document. Both HTTP and file URLs are allowed. Note that it is required to mention the protocol (http:// or file://) in the URL.
Using this feature, you can define custom code to be executed on the media platform in case of subsequent error events that would otherwise not be caught. Typically, this is code that deliberately does not lead back to the server, in scenarios where the error occurred due to connection problems.
The default is not to use a loop abort handler. If no location is defined in this field, the server will include an error handler holding a VoiceXML command that corresponds to the currently active dialog exit type. E.g., if the exit type is currently Disconnect, the server will render the following code into each VoiceXML document:
<catch event="error" count="10">
<disconnect/>
</catch>
i8 Note: If a Service object has been locked by another user, none of its settings can be modified. The Service object first needs to be unlocked.
For more information on the locking and unlocking of objects, refer to Chapter 7 Basic Commands in the Desktop for Eclipse Guide and Desktop for Web Guide.
Deploying a Service on a Server
In order for it to be available to a media platform, a service needs to be deployed on a server.
To deploy a service, open an existing Server object, add the service to the list of hosted services and save the Server object. For details on how to add elements to a list refer to Chapter 6 Object Editors in the Desktop for Eclipse Guide and Desktop for Web Guide.
In the Control Center select the corresponding server on the Server Manager tab (in Desktop for Web on the Server Management tab) and right-click it. From the context menu select Reload Service List. The new service will then be listed on the Service Manager tab (in Desktop for Web to be found in the Service Manager section on the Server Management tab).
In order to call into a service, it is necessary to also configure the media platform appropriately. For more information on this, refer to Media Platform Setup in Chapter 4 Service Deployment.