Cisco Agent Desktop Disconnects All Agents Simultaneously

This document describes one reason for Cisco Agent Desktop to disconnect all agents simultaneously from their respective sessions in a Cisco IP Contact Center (IPCC) Enterprise environment. This document also provides a possible workaround.

Note: For more information on the resolution of this issue, refer to Cisco bug ID CSCsa79930 (registered customers only) .

Prerequisites

Requirements

Cisco recommends that you have knowledge of these topics:

Cisco IPCC Enterprise Edition
Cisco Agent Desktop

Components Used

The information in this document is based on these software and hardware versions:

Cisco IPCC Enterprise Edition version 6.x and earlier
Cisco Agent Desktop version 6.x and earlier

The information in this document was created from the devices in a specific lab environment. All of the devices used in this document started with a cleared (default) configuration. If your network is live, make sure that you understand the potential impact of any command.

Conventions

Refer to Cisco Technical Tips Conventions for more information on document conventions.

Background Information

Contact centers organize agents into skill groups. Agents in each skill group are trained to handle specific types of calls. For example, literature requests are routed to skill groups in which agents are trained to provide general information services. Similarly, calls that involve complicated transactions are routed to more specialized skill groups.

Here is the general configuration in which this problem occurs:

Figure 1: IPCC Enterprise Edition

Problem

All agents configured to work in the contact center lose connectivity to the Cisco Agent Desktop at the same time. The agents include:

Agents who run Cisco Agent Desktop on their PC.
Agents who run Cisco Agent Desktop in a Citrix thin client environment.

When the mass disconnection occurs, agents need to reset the phones, and log back into Cisco Agent Desktop.

In addition to the mass disconnection, the Open Peripheral Controller (OPC) process on the Peripheral Gateway (PG) crashes with a Dr. Watson dump.

11:12:24 PG1A-opc Trace: Agent::GetSessionTimes - SkillGroup 40424 (0x9DE8)
priority=0 unknown for agent 1428 on Peripheral 5000.

Cause

When you review logs on the PG, check whether a skill group was removed. Here is the relevant section of the OPC trace log:

11:12:20 PG1A-opc UPDATED 1 SkillGroups. 
11:12:20 PG1A-opc UPDATED 1 Routes. 
11:12:20 PG1A-opc UPDATED 5 Agents. 
11:12:20 PG1A-opc REMOVED 1 SkillGroups.

Removal of a skill group causes the OPC process to crash. Here is the relevant section of the OPC trace log:

11:12:24 PG1A-opc Trace:
Agent::GetSessionTimes - SkillGroup 40424 (0x9DE8) priority=0
unknown for agent 1428 on Peripheral 5000.

The OPC process crash in turn causes the mass agent disconnection.

Solution

The best practice that Cisco recommends is to delete a skill group only when all agents who belong to that skill group are logged off.

Complete these steps in order to avoid the problem of mass disconnections from Cisco Agent Desktop:

Schedule a maintenance window when you want to delete skill groups.
Notify all the member agents in the skill group to log off from Cisco Agent Desktop.
Delete the skill group.
Notify the agents to log in again.