ACD Configuration

No changes are required to the actual Avaya ACD configuration beyond the changes mentioned in the Chapter 1:Avaya ACD Interface Requirements. However, some ACD-specific settings must be confirmed. This chapter describes these settings and provides guidelines that help you maintain your Avaya ACD and Unified ICM configurations.

Monitored VDNs and Inbound ACD Calls

A monitored VDN handles all inbound ACD calls initially. It is important that all VDNS involved in ICM call flow are monitored to ensure that there are no stale calls. A monitored VDN is equivalent to a configured Unified ICM Peripheral Target. For example, do not specify a Hunt Group Extension as the destination for inbound ACD calls. Hunt Groups that are vector-controlled (which is true for all skill groups in an EAS environment) cannot be monitored for calls.

The inability to monitor vector-controlled hunt groups is a restriction imposed by Avaya. An unmonitored call that reaches a Hunt Group or Agent cannot be tracked and accounted properly in Unified ICM contact or agent statistics.


Important

It is important that all VDNs to be monitored are properly configured as Peripheral Targets in the Unified ICM database.

Monitored Splits on CMS

The Avaya Hunt Group configuration screen for each monitored split on CMS must have its Measured field set to either "both" or "external." The field values are set in order for the CMS to receive Hunt Group (split) data.

Terminal Endpoint Identifier (TEI) Values

When you set up the Avaya ACD, the TEI value for an Avaya LAN or Avaya ACD is set to 1.

Configuring AES

Application Enablement Services (AES) software runs on an external server that communicates to Avaya Aura Communication Manager (or Avaya ACD) by using TCP/IP. It exposes a set of APIs that allow external applications like Cisco ICM to perform third-party call control and receive event notifications. The ICM PG uses either CVLAN or TSAPI link, which is a client/service software.

To best understand the configuration of the AES switch, begin with the Avaya documentation that shipped with your switch. The information provided here is meant to supplement but not replace the Avaya documentation.

We provide a limited amount of information to help you configure the switch to work with Cisco Avaya PG. The following tasks are described:

  • Setting Up the CV/LAN and TSAPI Links

  • Setting Up Agents and Hunt Group

  • Setting Up Call Routing

Setting up the CVLAN and TSAPI Links on AES Server

This section describes how to set up the CVLAN and TSAPI links on an AES server.

Perform the following procedure to establish the CVLAN link:

Procedure

Step 1

Open the AES OAM home page.

Step 2

Choose AE Services > CVLAN >CVLANlinks.

Step 3

On the CVLAN Link administration screen, click Add Link and perform the following:

  • Select the Signal

  • Uncheck the Proprietary check box

  • Select the Switch Connection

  • Select the Switch CTI Link Number

  • Select the CTI link version

  • Check the Heartbeat check box

Step 4

Click Apply Changes.

Figure 1. CVLAN Link Setup Screen


Adding CTI Client IP for a CVLAN Link:

  1. Open the AES OAM home page.

  2. Choose Administration > CTI Link Admin > CVLAN Links.

  3. Select the CVLAN link for which the client IP requires to be added and click Edit Client.

  4. Enter the IP address and click Add Client.

    Figure 2. Add CTI Client IP Screen


Step 5

Follow the procedure to establish the TSAPI link:

  1. Open the AES OAM home page.

  2. Choose AE Services > TSAPI > TSAPI Links.

  3. On the TSAPI Link administration screen, click Add Link and perform the following:

    • Select the Link.

    • Select the ACM (Avaya Communication Manager) to which you want to establish connection.

    • Select the Switch CTI Link Number.

    • Select the ASAI link version.


      Note
      The minimum link version for ASAI is 5.

    • Select the security as Unencrypted.


      Note
      Currently security enabled TSAPI Link is not supported.

  4. Click Apply Changes.

    Figure 3. TSAPI Link Setup Screen


  5. Select AE Services > TSAPI > TSAPI Links.

  6. Go to the TSAPI Link Properties section. Click Advanced Settings.

    Figure 4. TSAPI Advanced Settings


  7. Select Advertise only those Tlinks that are currently in service option as TSAPI Service Advertising Mode.

    Once the Tlink is created, navigate to the following path and note down the Tlink name. This is the name we use in PIM configuration.

    1. Open the AES OAM home page.

    2. Go to Security-> Security Database-> Tlink -> Tlink Name

Tlink Name for example:AVAYA#CMSIM#CSTA#AESSIM. This is applicable where:

  • AVAYA is a fixed constant.

  • Switch_Connection is a unique name, assigned to identify a switch (Communication Manager). In general, hostname of the switch is assigned as the name of Switch Connection in the AE Services server.

  • Service Type: refers to the CSTA service type. It can be either of the following:

    • CSTA — For using unencrypted TSAPI Link (non-secure connection).

      • CSTA-S — For using encrypted TSAPI Link (secure connection).

        The CSTA versus CSTA-S service types specify whether encryption is used between the application and the AE Services server.

  • AE_Services_Server_Name represents the hostname of the AE Services server which is assigned during the AE Services installation.

Adding CTI User in AES:

  1. Open theAES OAM home page.

  2. Go To UserManagement -> User Admin -> Add User Set the CT User option as "yes" and provide the remaining details as per Avaya documentation.


Note
A CTI user is unique for each PIM, which gets connected to an ACD.

Setting up Hunt Groups/Skill Groups

On the Avaya switch, a hunt group is a group of extensions to which similar calls are routed. A hunt group may include all agents who have a particular skill (for example, the ability to speak Spanish). It also includes all the agents who cover a geographical territory (for example, Boston sales). A hunt group is sometimes referred to as a skill group. By default, Avaya PG (TAESPIM) supports extensions of up to ten digits and this does not require registry configuration. The Agent Login IDs support up to ten digits.

Note

The Avaya PG (ECSPIM) supports extensions of up to ten digits – the agent can log in to a Softphone that has an extension up to ten digits. This ten-digit support applies to Agent Login IDs too. The Hunt Groups and VDNs support up to seven digits only.

In order to use a seven-digit, or ten digits, the config PIM registry EnableTenDigitExtension must be set to 1 in following path: 

HKEY_LOCAL_MACHINE\SOFTWARE\Cisco Systems, 
Inc.\ICM\<cus01>\<PGXX>\PG\CurrentVersion\PIMS\pim1\ATTData\Config\

If the registry EnableTenDigitExtension is set to 0, then it limits support up to five digits for extension, Agent Login IDs, Hunt Groups, and VDNs.

Some ACD systems provide a feature called EAS. For various reasons, you want certain agents to handle specific types of calls. For example, you require only your most experienced agents to handle your most important customers. You have multilingual agents who can serve callers in various languages. EAS allows you to classify agents according to their specific skills and then to rank them by ability or experience within each skill. Avaya uses these classifications to match each call with the best available agent.

Defining a Hunt Group/Skill Group for Agents

To set up agents, you must define a hunt group by completing the following steps:

Procedure

Step 1

Enter the command add hunt-group next and press Return. (You can also enter add hunt-group.xxx, where .xxx is the hunt group number.) The first Hunt Group screen appears.

Figure 5. Defining Agent Hunt Groups


Step 2

Complete screens 1 through 2 of the hunt group record as described in the Avaya documentation.

Step 3

Press Enter. The hunt group is successfully created.

Modifying the Agent Login ID

For each agent usingEnterprise Chat and Email, add the hunt group to the Agent login ID form.

To modify the Agent login ID, perform the following steps:

Procedure

Step 1

Enter the command change <agent login ID number>. The Agent Login ID screen appears.

Figure 6. Modifying Agent Login ID


Step 2

Add the hunt group that indicates Unified CCE agents in the SN (Skill Number) field in the table at the bottom of the screen.

Step 3

Complete the remaining fields as described in the Avaya documentation.

Setting up Call Routing

After you have set up your Enterprise Chat and Email agents and the phantom lines, ensure that the Avaya switch routes calls to them appropriately by:

  • Writing a vector to route calls

  • Creating a VDN to access the vector

Write a Vector to Route Call

A vector is a set of instructions the switch follows to ensure that the right call gets to the right agent. Whether you use predictive or phantom CTI strategies, write a vector that routes appropriate incoming calls to an Enterprise Chat and Email agent hunt group. Write a vector for each group to which you want to route calls, and you require the hunt group number established for Enterprise Chat and Email agents when setting up a vector to route calls to those agents.

To create a vector:

  1. Enter the command change vector xx (where xx is the vector number) and press Return. The Call Vector form appears.

    Figure 7. Vector


  2. Complete the Call Vector screens 1 through 6. Instructions for completing these screens are provided in the Avaya documentation.

Example of a Post route Vector for a Call

Following is an example of a Post route call vector. This sends adjunct route request to the Unified CCE Avaya PG which is connected on routing link 1.

Figure 8. Call Vector


Create a Vector Directory Number

After setting up a vector for Enterprise Chat and Email calls, set up Vector Directory Numbers (VDNs) to direct incoming calls to that vector. You can create several VDNs that refer to the same vector, ensuring that calls from a variety of sources can be routed to the same skill group.

To create a VDN, complete the following steps:

Procedure

Step 1

Enter the command add VDN.xxxx (where xxxx indicates the VDN). The Vector Directory Number screen appears.

Figure 9. GUI for Vector Directory Number


Step 2

Complete the screens 1 through 3 based on the instructions in the Avaya documentation.

Configuring Return Destination VDN on Avaya Switch

The Return Destination automatically redirects a call from one monitored VDN to another VDN for continuous call processing, after an agent disconnects the call.

If the call flow involves two post-route VDNs, the call variables are preserved from the first route request to the second route request when call gets automatically redirected from a post-route VDN to another post-route VDN for continued call processing (after an agent disconnects from the call). (This is according to the return destination feature enabled on Avaya ACD.)

If the call flow involves a first post-route VDN and then a non post-route VDN after return destination, the call variables are still preserved from the first call (post-route request) to the second call after return destination.

After return destination, the last agent also has the call variables which were set in the first call prior to return destination.

The Avaya PIM detects whether the return destination is configured on a VDN, by verifying the parameter string of that VDN in the Peripheral Monitor tab of the PG Explorer. The PIM then sends a NEW_TRASACTION_IND to OPC, allowing the OPC to preserve the call variables in the second route select.

For example:

  1. VDN 32222 is the first post route VDN. VDN 32223 is the label returned for 32222. It is the Return Destination VDN as well.

  2. After Return Destination occurs on VDN 32223, the call variables are preserved in the second post route. VDN 32224 is the Post Route VDN. It is configured as Return Destination VDN for VDN 32223 on Avaya Switch.

    When Agent on VDN 32223 drops the call, Return Destination occurs and call is redirected to VDN 32224. A NEW_TRANSACTION Indication is sent to OPC.

The Return Destination can be configured on the Avaya Switch. The following example explains the configuration of this feature:

Procedure

Step 1

Click Tab 1 on the VECTOR DIRECTORY NUMBER screen and set the following field:

  1. Allow VDN Override: Set this field as y.

Step 2

Click Tab 2 on the VECTOR DIRECTORY NUMBER screen to set the return destination VDN for 3606. Set the Return Destination field as 3001. This value is configured as return destination VDN for 3606. When an agent on VDN 3606 drops the call, it is automatically redirected to VDN 3001.

To set up Return Destination VDN on Unified ICM, see the section Configuring the Return Destination on Unified ICM

See the section ACD Notes and Restrictions for known caveats for Return Destination VDN.

Ethernet Busy Hour Call Rates

Each Avaya Ethernet CTI link can support a BHCR of approximately 32,000 in general use by the PG (that is, without Post-Routing or third-party call control). This value is an approximation and may be affected by the number of agents, anticipated peak busy hour call rate, average number of CTI events/calls, and the number of splits, trunk groups, and VDNs. Provision a dedicated Ethernet CTI link for Unified ICM application.

Post-Routing, Station Monitoring, Third-Party Call Control

If Post-Routing, station monitoring, or third-party call control is performed on the same Ethernet CTI link, the link supports up to 20,000 BHCR due to additional message traffic. Depending on your configuration, you might need an additional Ethernet CTI link to be used exclusively for Post‑Routing, station monitoring, or third-party call control.

An Ethernet CTI link dedicated exclusively to Post-Routing (that is, no event monitoring) can handle approximately 64,000 BHCR. Calculating throughputs for third-party use is dependent upon the number of stations involved and anticipated usage. In general, third-party usage on the CTI link uses some of the CTI bandwidth.

Active Association Limit

Active associations are used for all requests made of the switch. Some of the requests made of the switch remain open for an indefinite period of time (for example, event notification requests, monitoring VDNs). Other requests end when the switch returns the response (e.g., value query for time-of-day). The indefinite requests include VDN event monitoring, station monitoring, and skill group monitoring.

ASAI_TEST utility

ASAI_TEST is a utility that allows you to check the connectivity between Unified ICM PG and the Avaya ACD (Avaya ACD can include either Avaya ACD card or AES Server).

Before running the ASAI_TEST, ensure the IP connectivity between the PG and the Avaya ACD card. To do so, initiate a ping test from the PG to the Avaya ACD card. If the ping test passes, you can proceed with the ASAI_TEST.

To run ASAI_TEST, use this command syntax: 

<Directory>:\icr\bin>asai_test
usage: asai_test [-m hostname/IP address] node_id

Note
The node_id is also referred to as the CTI link number. The maximum number of CTI links can be 8.

Note
The Active Association Limit is not applicable to TAESPIM.

Maximum Agent and BHCA

Unified ICM software (CC, PG, CTI server) currently supports 3000 Agents and 60000 BHCA.

Set the dynamic PIM registry "BriMaxOutstandingMessages" to 100 (decimal) in the following path: 

HKEY_LOCAL_MACHINE\SOFTWARE\GeoTel\ICR\<cus01>\<PG1A>\PG\CurrentVersion\PIMS\pim1\ATTData\Dynamic

Call Handling Methods to Avoid

Following are the call handling methods you need to avoid:

  1. Avoid setting up station coverage paths where all internal calls are marked to go to coverage.

  2. Avoid having agents transfer calls directly to other agent stations or to other agent IDs. Instead, the calls can be transferred to a hunt group (or split). The calls can also be transferred to a VDN to ensure proper call monitoring.

Universal Call ID

If the Avaya Universal Call ID (UCID) is preferred. The field Send UCID to ASAI is set to Yes. You can do this through Feature-Related System Parameters form on the Avaya.

Starting with ICM 7.5(9), the Avaya PIM is enhanced to use UCID information from the Avaya Switch to clean up old calls in the Avaya PIM.

To enable the UCID feature on the Avaya Switch, change the following system-parameters features:

  • Create Universal Call ID(UCID) ? y

  • What is the value for UCID Network Node ID? <any value_unique to a switch>

  • Is it necessary to send UCID to ASAI ? y

The UCID value is stored in the Unified CCE central database, in the Termination Call Detail table, in the CallReferenceID field. For more information, see the Database Schema Handbook for Cisco Unified Contact Center Enterprise.

CVLAN Link Configuration

CVLAN link configuration is set to have Event Minimization set to Yes. This is especially important if you are using third-party functionality.

To set Event Minimization to Yes:
  1. Stop the PG.
  2. Busy-out and release the CVLAN link. This activates this CVLAN link attribute.
  3. Restart the PG.
When Event Minimization is enabled, the Unified ICM PG CVLAN links are dedicated to the PG (that is, no other applications are using those CVLAN links). In a duplexed environment, both PG sides can use the same CVLAN links.

CMS Cisco Real-Time Report

The guidelines in this chapter are intended for Avaya installers of the CMS Cisco RTA report:

  • Skillnums argument: The CMS report uses the skillnums argument. The Unified ICM PG started supporting skillnums in Unified ICM software. Therefore, any installation of Unified ICM software uses skillnums. This applies to all the new installations and upgrades to ICM software.

  • Agent login: For CMS report version 3.5 or later, the PIM does not log agents into a skill group unless the skill group is monitored by CMS. This is a requirement because CMS does not pass agent state data or logout events for non-monitored skill groups. The lack of this CMS data can cause agent count and agent state mismatches. Conversely, if the CMS report is a pre-3.5 version, the PIM still logs agents into all groups provided in the agent login event, but CMS does not provide logout events for the non-monitored skill groups.
Because no logout record (or any agent state record, for that matter) is provided by CMS for these non-monitored skills, and because the version of the CMS report is pre-3.5, the PIM may leave agents in their last state. For this reason, use a CMS report that is version 3.5 or later.

Note
The Avaya PG currently supports 20 skills per agent. The enhanced RTA 5.0.5, which supports 60 skills per agent, is not supported by Unified ICM.

  • Noskillnum flag Make sure that the noskillnum flag is set to skillnums (that is, CMS provides the list of monitored skills) in the following list of files. Split/Skill numbers need to be in the CMS startup header provided to the PG.

    The following files are impacted:

    • Startrta

    • testrta

    • skills1

    • skills2, and so on

These are files on the CMS machine.

  • Multiple CMS reports on one PIM If multiple CMS reports are configured for a single PIM on a PG, the CMS report must use the timestamp argument. The timestamp argument causes the CMS report to include a UNIX timestamp in each of the records sent to the PG. The PG requires the timestamp to properly order the incoming CMS records from the multiple reports.

  • Agent-skill pairs Upgrade to the newest CMS report if you find you need increased agent-skill pairs. The newer CMS reports can be configured to support up to 10,000 agent-skill pairs (default 2,400). Using this single (increased) agent-skill pair capability eliminates any need for using multiple CMS reports (and therefore not require timestamps in the CMS reports).


Note
In later CMS reports, the arguments (for example, noskillnums ) described above may have changed. Therefore, Avaya installers check for the correct arguments to achieve the desired functionality as described above.

CMS Minimum Refresh Rate

The CMS report is installed to run as an administrator in order to allow a minimum refresh rate of three seconds. It is necessary for you to ensure that the refresh rate used for the custom CMS report is allowed for an administrator CMS login. In addition, administer the CMS report via the appropriate login (for example, CMS). Using another login to administer the report does not work. The Avaya Professional Services group can provide the details on which login are used to administer the report.

If agents are being dynamically re-skilled (logged into and out of skill groups with some frequency), it is possible that the CMS report does not see an agent logout/login sequence for a skill group. For example, if the agent is logged into skill 1 and is logged out of and back into skill 1 within the CMS refresh rate period (that is, in between CMS snapshots of data), then CMS does not see this logout/login transition.

Dynamic re-skilling is supported only when the agent is in "Available" or "AUX" state. If the re-skilling is attempted for an agent who is in a state other than these states, then the reskill record is held by CMS until the agent state changes back to “Available” or “AUX” state.

Configuring the CMS Report

While configuring the CMS report the following data items are required:

  • ACD Number: The ACD number is the Avaya ACD number as known to the CMS system.

  • Refresh Rate: The refresh rate is the rate at which the report captures the agent data and pass it on to the PG. The minimum refresh rate is 3 seconds. A typical refresh rate is 10 seconds.

  • Splits to Monitor: The splits to monitor are the ACD skill groups, that are required to be monitored by the PG. It is mandatory to update the list of CMS monitored splits periodically.

  • Avaya type

    For example, Non-EAS or EAS.

  • PG LAN Information This includes IP address, netmask, and hostname. The CMS and PG are usually on the same LAN. The PG LAN information is required by the Avaya Professional Services engineer to set up the CMS report, so that it is connected to the PG. After the initial installation, this is never changed.

Important
If any of the above CMS information is changed (for example: Monitored splits or the refresh rate), the CMS report and PG must be stopped and restarted in order for the changes to take place.

Avaya Configuration for “CMS-less” PGs

In a PG configuration that does not use CMS, additional configuration is necessary on Avaya.

  • PG requires skill groups to be monitored to track agent login and logout events. No agents can log in to that skill group if a skill group is not monitored. PG uses 3PDC or Monitor request APIs to monitor a skill group, based on interface.

Avaya currently restricts one application to third-party domain control of a skill group.

  • Enable Event Minimization for CVLAN CTI links used by the Peripheral Gateway. This is not applicable when PG uses TSAPI Interface to connect to AES.

  • For optimal performance, external applications that alter agent state on the Avaya ECS, use the Enterprise CTI interface. Contact your Cisco Unified ICM software representative for comprehensive and up-to-date information on configurations.

ACD Notes and Restrictions

Following are the notes and restrictions applicable to Avaya:

  • Monitoring VDNs - It is important that all VDNs to be monitored are properly configured as Peripheral Targets in the Unified ICM database.

  • CTI links and BHCC- Each CTI link can support the following link specification (approximate):

    • 8,000 BHCC using a BRI CTI link

    • 32,000 BHCC on a G3r using an Ethernet CTI link

    • 20,000 BHCC on a G3i or G3s using an Ethernet CTI link

    These link specifications are derived from Avaya-provided data and are subject to change.
  • The Avaya PG supports Agent IDs, Agent Extensions, Hunt Group Extensions, and VDNs that start with a zero. This is supported for both CMS and CMS-less environments.
  • The Avaya PG does not support Hunt Group Numbers that start with a leading zero.

  • Intermittent Failure of Network Transfer: There is occasionally a timing issue in the set of events involved in a Network Transfer. Due to this issueIntermittent Failure of Network Transfer, the NIC Call ID is not populated in the transfer call. This results in call failure. To avoid this problem, introduce a delay of 1 second in the Vector for the post-route number (VDN); that is,

Due to this issueIntermittent Failure of

01 wait-time 1-secs hearing silence

02 adjunct routing link 1

  • Avaya ECS PIM Failure - The ECS PIM stops functioning when the AES link on the Avaya switch sends a "Busyout" command.

  • VDN Return Destination

    • The Return Destination VDN feature does not support the call conference before agent drops the call (expecting return destination to set in). That is, the PG loses its track of the VDN to which the call was originally delivered and call variables are not retained post Return Destination. However, this feature supports call transfer before the agent drops the call (expecting return destination to set in). That is, the PG tracks the VDN to which the call was originally delivered and call variables are retained post return destination.

    • Return Destination cannot be run multiple times. That is, Return Destination can occur only once for a call in Avaya.

      See sections Configuring Return Destination VDN on Avaya Switch for Configuring the Return Destination VDN on Unified ICM respectively.

  • Processing Invalid CMS Records - Whenever the CMS receives an invalid record, the Avaya PG service goes out of service. Create the following registry value and set its registry value data as 1 so that Avaya PG Service ignores the invalid CMS record and continue its usual operation.

    • Registry Value: IgnoreInvalidCMSRecord

    • Registry Value Type: DWORD

    • Registry Value Path: HKEY_LOCAL_MACHINE\SOFTWARE\Cisco Systems, Inc.\ICM\<cust_inst>\<pg_inst>\PG\CurrentVersion\PIMS\<pim_inst>\ATTData\Dynamic\

    • Registry Value Data:

      • 0 – Avaya PG Service will go out of service and log out all the Agents, if an invalid CMS record is received.

      • 1 – Avaya PG Service will ignore the invalid CMS record and continue its usual operation.

Multiple PGs

The Avaya ACD allows connections from multiple PGs. However, while using such a configuration, the resources (like Stations, Agents, VDNs, Splits, and any other resources) used by each PG, are maintained as separate configurations.

Multiple PG deployments on a single ACD are used to split the load on the PG. These can also have a dedicated PG to service a business line in the contact center.

To deploy multiple PGs on a single ACD, it is required for you to have distinct configuration between the PGs. Section Dual PG Setup describes the configuration of two PGs on a single ACD. You can follow the same steps to configure multiple PGs to the same ACD.


Note

Contact the ACD vendor for ACD-related issues or limitations on connecting multiple PGs to a single ACD.

Dual PG Setup


Note
Ensure that you adhere to the requirements provided in this section while deploying multiple PGs. The performance and functionality of the PG is impacted if you do not follow the requirements listed in this section. When two PGs are deployed, ICM routing effectively sees the Avaya system as two independent peripherals. As such, there are some specific configuration and operational requirements that are required to be put in place. The following figure describes the overview of Dual PG setup.
Figure 10. Dual PG Overview


For the same Avaya ACD to behave as an independent peripheral to Unified ICM, avoid having one skill group monitored by the two PGs. To achieve this, you have to determine which Avaya skill numbers are associated with PG1 and PG2, respectively. For example, skill groups 1-1000 and 1001-2000 would be two separate sets monitored by each instance.

When two PGs share the load of a single PG, have a logical correlation between the skill numbers associated with the different PGs. For example, the "pre-paid sales" skill associated with PG1 might be 500, and the one associated with PG2, 1500 (just added the digit 1 in front). It is essential for the supervisor to look at a report that combines the information from the two corresponding skills to understand the overall skill performance. After the correlation is defined, each agents are assigned only the skill numbers. These numbers belong to the same PG (considering the example stated before, an agent must not have skills 500 and 1502 at the same time).

Because the PGs also monitor Avaya stations, the agents associated with that PG must log in to the stations monitored by that PG only. All stations and agent-IDs at a given physical site is required to be defined at only one PG. You can avoid having one site with entities from two PGs. Assign stations sequentially for every PG.

The VDNs (that are monitored by Unified ICM) are independent, regardless of whether they are used by calls when they first enter the environment or for translation routes. The CTI links used by these VDNs (through the vectors they point to), are separated; the CTI links established with PG1 A and B, and with PG2 A and B. You can also define dial-plan ranges for VDNs in each PG to make the configuration simpler (but it is not required).

The calls within the PG can be dialed directly using usual dial plan numbers such as VDNs, Agent IDs, extensions, and hunt groups. Ideally, the calls are not disconnected across two PGs. However, in such scenarios, the call has to be translation routed to the target PG over trunk, which is equivalent to routing to a PG connected to a different switch. To achieve this kind of a routing, loop back trunks is required to be provisioned on the switch and used for routing calls to dial plan numbers of another PG.

When the call is routed from PG1 to PG2, the target PG2 understands the call as an inbound call and the ICM reporting reflects the same. To prevent inappropriate agent behavior, the Avaya system can be programmed to block incorrect call flows (COR or tenant settings). Design the system to avoid or minimize the call between the two PG groups.

Cisco mandates that trunk groups monitored by each PG be separate. If two PGs are used to monitor the same trunk group, Unified ICM software do not understand that the feeds, it gets are duplicated.

Other Avaya resources such as announcements, classes of restriction, CMS reporting, recording, and so on, are not affected with dual PG implementation.

Note
Although two PGs can provide scalability from a Cisco Unified ICM perspective, it is also necessary for you to consult Avaya about how ACDs handle the increased CTI traffic. It also considers all other applications that currently used, such as recording and virtual hold. These thresholds are associated with a large number of CTI-enabled agents in the Avaya ACDs, although, two PGs are being used.

Maintaining Your Configuration

It is preferred that changes made to your configuration are accomplished on the Avaya/CMS and in the Unified ICM database consecutively. This ensures that the PG gets the configuration updates on the Avaya/CMS systems.

It is imperative that the Avaya, CMS, and the Unified ICM database configurations are kept synchronized (that is, up-to-date with each other). Inaccurate or incomplete data results in inaccurate agent or call data.

Configuring High Availability CMS

The high availability CMS configuration minimizes the down time in the event of Avaya CMS failure. If you want to have such configuration in your call center environment, set up the PGs as follows:

  • Duplex PG configuration: Both CMSs (CMS no.1 and CMS no.2) require to use both PG's IP addresses (that is, IP addresses of PG-A and PG-B) for connections. It is the same as they do in a single CMS configuration. However, at any given time, only one CMS can connect to the active PG. The other PG is always in standby mode. So if PG-B is currently active, PG-A is in standby mode (and conversely).

    When running the Peripheral Gateway Setup tool, ensure that the CMS Hostname field in the AvayaPIM Configuration pop-up menu is blank. This allows the PIM to accept a connection from either HA CMS server. If one HA CMS server goes down, the other initiates a connection to the PIM on the active PG.


Note
Irrespective of this being a High Availability CMS configuration or not, a CMS Data Feed failure results in a failover from one side of a duplexed PG to the other.
  • Simplex PG configuration: Both CMSs use the PG's IP address for connection. But only one CMS can connect to the PG at any given time. Also, be sure when running Peripheral Gateway Setup tool that the CMS Hostname field in the Avaya PIM Configuration pop-up menu is blank. This allows the PIM to accept a connection from any one CMS server. If one CMS server goes down, the other initiates a connection to the PIM.

Note
The HA CMS server going down, leads to CMS Data Feed failure; this results in a brief outage on a simplex PG.