Planning Your Deployment

[Modified: 8.5.102.06, 8.5.141.04]

Before you deploy Workspace, you should take time to define your needs in terms of load, bandwidth, scale, the type of network that you have or want to develop, the number of resources you plan to manage, and the type of deployment that you want:

• ClickOnce
• Non-ClickOnce

Defining your needs

This section provides items that you should consider when you are planning your deployment.

Load, IIS vs. Apache

Workspace is designed to be equally compatible with Microsoft Internet Information Services (IIS) or Apache web servers. Your choice depends on the server-side operating system and HTTP server that you are running. Refer to the following system guides for details on compatibility and system requirements:

• Hardware Sizing
• Genesys Interoperability Guide
• Genesys Supported Operating Environment Reference Manual

Type of network

Refer to the following system guides for details on compatibility and system requirements:

• Hardware Sizing
• Genesys Interoperability Guide
• Genesys Supported Operating Environment Reference Manual

Single Sign-On (SSO)

[Added: 8.5.102.06]

Workspace supports Single Sign-On authentication by using Kerberos User Authentication. This feature must be configured prior to deployment.

Choosing between a ClickOnce deployment and a Non-ClickOnce deployment

ClickOnce A ClickOnce deployment of Workspace automatically handles software updates as you make them available in your environment. If you do not have the ability to push applications, updates, and configurations to your agents, you might want to take advantage of a managed services deployment approach in your environment by using ClickOnce.

Non-ClickOnce In a Standard Deployment, you must install the Workspace application on each client workstation. In this scenario, you must push software updates to each workstation in your environment. Refer to the following sections for information about different deployment scenarios:

• Deployment Overview
• ClickOnce Deployment
• Non-ClickOnce Deployment

Memory usage

This Table represents the Memory Usage range of Interaction Workspace. Minimum value is the out of the box version using voice only interactions Maximum value is the out of the box version using multimedia interactions

OS Type	Memory used
x86 (32-bits)	180 - 280 MB
x64 (64-bits)	200 - 300 MB

Interaction Workspace can use more memory if deployed with click-once on compatibility mode with 8.1.2. This mode runs Interaction Workspace in 64-bits native mode and can use up to 450 MB.

If you are using Workspace SIP Endpoint, an additional 60 MB are used.

Monitor resolution

The minimum supported resolution is 1024 x 768 at 100% appearance display factor.

Effects of configuration options and privileges on performance

These table list the effects that some configuration options and privileges might have on network bandwidth and also the performance of Configuration Server and Configuration Server proxies.

Summary of the effects of Workspace options and privileges on network bandwidth

Option/Privilege	Default Value	Values that might affect system network bandwidth	Functional impact of using different values
Workbins - Can Use My Team Workbins	Unassigned	Assigned During supervisor/Team Lead login, the Workbin module loads the current state of each of the workbins of the agents on the supervisor's team to provide the supervisor with an overview of the content of each workbin. This action generates a set of requests to Interation Server. The bandwidth that is consumed by those requests is proportional to the following variables: the number of monitored agents the number of workbins assigned the supervisor the number of interactions in each workbin the size of the interaction properties in each interaction (depends on the Business Process design	No Team Workbin Supervision
teamcommunicator.list-filter-showing	Agent	Agent Affected instance: Configuration Server (Proxy)	Target types are not displayed to agents
login.enable-place-completion	true	true: Workspace loads all the Places that are visible to the logged-in agent immediately to enable the process of Place. This might be a large number of Places in a large scale environment. Affected instance: Configuration Server (Proxy)	false: agents must enter their Place name manually and the verification is performed after the Place is submitted. If a default place is assigned to agents, this issue is mitigated.
teamcommunicator.load-at-startup	true	true: all configured object lists for team communicator are loaded during agent login and added to the index. Affected instance: Configuration Server (Proxy)	false: all configured object lists for team communicator are loaded at the first time that the Team Communicator is used.
general.configuration-object-collection-cache-timeout	0	0: No local cache of the list of objects is maintained; therefore, every time that the Team Communicator is initialized, the list of objects is requested from Configuration Server.	Any positive integer: Specifies the number of hours between requests for objects from Configuration Server, reducing the number of requests made to Configuration Server.
general.configuration-agent-collection-loading-method	read-objects	read-objects: This is the legacy agent-retrieval method. This method returns the list of agents according to the "read" permission of the current agent. For each retrieved agent this method returns the full agent data from Configuration Server, including annex structure, in an uncompressed format. The bandwidth consumed can be much larger than the other retrieval methods. When an agent logs in to Workspace, several collections of configuration objects are loaded from Configuration Server or Configuration Server Proxy to build a Lucene index which enables instant quick search in Team Communicator. Depending on the user profile, this mechanism triggers the download of different types of objects, such as: agents, DNs (Routing Point, Queues), Scripts (Interaction Queue, Workbins), Agent Groups, and/or Skills. In large scale environments where Workspace is configured to enable searching of agents in Team Communicator, Workspace might cause a large load on Configuration Server/Configuration Server Proxy, and on the network. This is particularly true of massive concurrent logins, such as at the beginning of a shift or after an infrastructure incident.	brief-info: For each retrieved agent, this method returns only a subset of the agent data. This can result in significant network bandwidth optimization, as well as a decrease of the load on Configuration Server. The larger the contact center, the larger the size of the Person collection. Each person object returned by Configuration Server Proxy can be a large data set that contains annex KVCollection where Workspace can store the agent profile and where other custom business applications can also store data. This can result in several megabytes of download for each agent login.

Summary of Interaction Workspace options and privileges that can affect Configuration Server (or CS Proxies)

Option/Privilege	Default Value	Values that can affect Configuration Server (or CS Proxies)	Functional impact of using different values
teamcommunicator.person-cache-for-favorites-recents-enabled	false	false (default): represents the legacy behavior where the preparation of Favorites and Recents can generate some duplicated requests to Configuration Server.	true: optimizes the usage of local cache when loading Recents and Favorites. The Team Communicator user experience during data loading can be slightly different when this option is set to 'true'.
options.record-option-locally-only	false	false and the options.record-location option is absent or left empty: the agent profile is stored in the annex of the corresponding Person object when an agent logs out of the application. Write requests are transmitted by Configuration Server Proxies back to the central Configuration Server and then the Central Configuration Server notifies all proxy instances about the update. Affected instance: central Configuration Server and Configuration Server Proxies.	false and options.record-location set to a valid shared directory: there is no functional impact. Refer to Storing the agent profile on a controlled shared host. true: Unless you are using the Windows Roaming Profile approach as part of your IT policy, the personal settings do not follow an agent who is roaming or hot seating.
general.configuration-update-notification	All	All or <empty>: Workspace subscribes for notifications about all object types that are read. The Agent option might also generate a lot of notifications, depending on Configuration Server operations. Affected instance: Configuration Server (Proxy)	None: no notification at all. Any config update is taken into account at next login. ThisApplication, ThisAgent: Workspace is informed about modifications to the configuration of the current agent or current Application. Any other changes are taken into account at the next login.
teamcommunicator.list-filter-showing	Agent	Agent Affected instance: Configuration Server (Proxy)	Target types are not presented to agent
login.enable-place-completion	true	true: Workspace loads all the Places that are visible to the logged-in agent immediately to enable the process of Place. This might be a large number of Places in a large scale environment. Affected instance: Configuration Server (Proxy)	false: agents must enter their Place name manually and the verification is performed after the Place is submitted. If a default place is assigned to agents, this issue is mitigated.
interaction.evaluate-real-party-for-agent display-format.agent-name	true	true: Workspace accesses Configuration Server and Stat Server before an interactive notification is displayed, to retrieve the display name of internal agents or supervisors who are engaged. The generated load on Config Server Proxy is not large, but it is proportional to the flow of interactions. Use the interaction.evaluate-real-party-for-agent.expression option to use a regular expression to display an agent's name instead of the agent's DN. Affected instance: Workspace client	false: the internal voice interaction parties are displayed as phone numbers instead of a display name.
teamcommunicator.load-at-startup	true	true: all lists of configured objects for Team Communicator are loaded at login time and added to the index, which can affect the system in a scenario where there is massively concurrent agent login operations. Affected instance: Configuration Server (Proxy)	false: all lists of configured objects for Team Communicator are loaded the first time that Team Communicator is used, which might make the first activation of Team Communicator slower.
interaction.override-option-key interaction.disposition.value-business-attribute interaction.case-data.format-business-attribute	empty	Business Attributes and Transaction objects for Dispositions and Case Data are always loaded the first time that an interaction requiring them is received by an agent. They are then cached for future use. The more possible values that exist, the more accesses are required during interaction notification. The generated load is not large, but in case of slow response time, there might be a delay before the interactive notification is displayed. Affected instance: Workspace client	The general.configuration-business-attribute-cache-preload and general.configuration-business-attribute-folder-cache-preload options enable you to cache a list of Business Attributes up front to avoid a slow response accessing the Business Attributes in scenarios where Config Server has difficulties answering in a timely manner. [Modified: 8.5.141.04]