Administering Queues, Logs, and Tasks

This chapter discusses how to:

Administer physical queues.
View queue server, queue, and agent state.
View broadcast, chat, and event logs.
Administer overflow and escalated tasks.

Administering Physical Queues

To administer physical queues, use the Physical Queues (MCF_ACCPT_TASK_CMP) component.

This section discusses how to:

Move agents between physical queues.
Move queues.
Balance queues.

Pages Used to Administer Physical Queues

Page Name	Object Name	Navigation	Usage
Move agent	MCF_DEMO_AGENTQ	PeopleTools, MultiChannel Framework, Universal Queue, Administration, Physical Queues, Move agent	Move an agent to a new physical queue.
Move queue	MCF_DEMO_QUEUEQ	PeopleTools, MultiChannel Framework, Universal Queue, Administration, Physical Queues, Move queue	Move agents and tasks from one physical queue to another on the same logical queue.
Balance queue	MCF_DEMO_Q_BALANCE	PeopleTools, MultiChannel Framework, Universal Queue, Administration, Physical Queues, Balance queue	Balance agent and task assignments over active physical queues.

Moving Agents Between Physical Queues

Access the Move agent page.

You can move an agent, and any open persistent tasks that are associated with that agent, from one physical queue in one cluster to another physical queue in another cluster on the same logical queue. For example, if an agent is transferred to another office, you can move the agent to another physical queue on a server closer to the new office.

Logical Queue	Select the logical queue within which the agent is to be moved.
Number of Accepted Persistent Tasks	Displays the number of persistent tasks this agent has accepted on this physical queue. This number is updated and displayed when you select the logical queue.
Refresh number of accepted tasks	Click to update the number of persistent tasks accepted by this agent on the current physical queue.
New Physical Queue	Select the new physical queue to which this agent and the persistent tasks accepted by this agent will be assigned.
Move agent to new physical queue	Click to perform the action.

Note. Ongoing chat sessions, which are not persistent tasks, are not affected by the move agent action.

Moving Queues

Access the Move queue page.

You can move all agents and their open persistent tasks from one physical queue to another physical queue within the same logical queue. For example, to delete a physical queue, move its agents and persistent tasks to another queue before deleting the first queue. Or, if the cluster serving this physical queue is overloaded, you can create another physical queue on another cluster, mark the first physical queue as inactive, create another physical queue on another cluster and move agents and persistent tasks from the inactive physical queue to the new physical queue.

Logical Queue	Select the logical queue within which the selected physical queue's agents and persistent tasks will be moved.
Physical Queue	Select the physical queue from which agents and persistent tasks will be moved.
Number of accepted tasks and Number of assigned tasks	Displays the number of accepted and assigned tasks open on this physical queue.
Refresh number of tasks and agents	Click to update the number of agents and persistent tasks assigned to this queue.
Number of enqueued tasks and Number of agents	Displays the number of enqueued tasks and assigned agents on this physical queue.
To Physical Queue	Select the physical queue to which the currently assigned agents and tasks will be moved.
Move agents and tasks	Click to move the assigned agents and tasks to the specified physical queue. Note. The physical queue must be inactive before moving agents and tasks. Inactivate the physical queue on the Queues page. Agents on the inactive physical queue are automatically logged off before being moved.

Balancing Queues

Access the Balance queue page.

Over time, the distribution of agents and their associated skill levels, languages, and so on, across the physical queues belonging to a logical queue may change as agents are added or deleted. Rather than manually rebalancing the queue by moving individual agents, you can use the Balance queue page to randomly reassign agents and their open persistent tasks across all the active physical queues belonging to the selected logical queue.

Logical Queue

Select the logical queue across which agents and persistent tasks will be balanced.

Randomly reassign agents and tasks on active physical queues

Click to balance agents and tasks across the active physical queues for the selected logical queue.

This action redistributes agents and tasks assigned to this logical queue across all active physical queues without regard to previous assignments. Ensure that agents assigned to the affected physical queues have shut down their MultiChannel Consoles. They are not logged off automatically.

Viewing Queue Server, Queue, and Agent States

To view queue server state, queue state, and agent state, use the Queue Server State (MCF_QSERVSTATE_CMP), Queue State Summary (MCF_QUEUESTATE_CMP), and Agent State Summary (MCF_AGENTSTATE_CMP) components.

This section discusses how to:

View the queue server state.
View the queue state summary.
View the agent state summary.

Pages Used to View Queue Server, Queue, and Agent States

Page Name	Object Name	Navigation	Usage
Queue Server State	MCFQSERVSTATE_PG	PeopleTools, MultiChannel Framework, Universal Queue, Administration, Queue Server State	Examine the state of a queue server at a particular time.
Queue State Summary	MCFQUEUESTATE_PG	PeopleTools, MultiChannel Framework, Universal Queue, Administration, Queue State Summary	Display the state of a queue.
Agent State Summary	MCFAGENTSTATE_PG	PeopleTools, MultiChannel Framework, Universal Queue, Administration, Agent State Summary	View the agent state.

Viewing the Queue Server State

Access the Queue Server State page.

The Queue Server State page displays cumulative diagnostic totals for the selected queue server, at the selected time, across all physical queues serviced by this queue server. Cumulative totals are reset after each state dump.

This state information comprises statistics to measure load and throughput that can be analyzed and used for tuning the system. For example, if counts are very high, consider adding another physical queue to balance the load.

The state stamps are inserted for every primary queue server by cluster at configurable intervals, irrespective of user activity. The intervals are configured on the Cluster Tuning page.

Viewing the Queue State Summary

Access the Queue State Summary page.

The Queue State Summary page displays cumulative diagnostic totals for the selected physical queue.

This state information comprises statistics to measure load and throughput that can be analyzed and used for tuning the system. For example, if counts are very high for this queue, consider adding another physical queue to balance the load.

The state stamps are inserted for every physical queue at configurable intervals, irrespective of user activity. The intervals are configured on the Cluster Tuning page.

Viewing the Agent State Summary

Access the Agent State Summary page.

The Agent State Summary page displays cumulative totals for the selected agent.

This state information comprises statistics to measure agent performance and status that can be analyzed and used for performance evaluation.

The state stamps are inserted for every agent at configurable intervals, irrespective of user activity. State is only recorded for agents currently logged on. The intervals are configured on the Cluster Tuning page by setting the dumpinterval. Enable agent logging by setting dumpagent to yes.

An agent state of Active indicates the agent is available; Inactive indicates the agent is not available.

Note. On a busy system, recording state information frequently may affect performance.

Viewing Broadcast, Chat, and Event Logs

To view broadcast, chat and event logs, use the Broadcast Log (MCF_BCAST_LOG_CMP), Chat Log (MCF_CHAT_LOG_CMP), and Event Log (MCF_EVENTLOGL_CMP) components.

This section discusses how to:

View broadcast logs.
View chat logs.
View event logs.
View MCF logs.

Pages Used to View Broadcast, Chat, and Event Logs

Page Name	Object Name	Navigation	Usage
Broadcast	MCF_BCASTLOG_PG	PeopleTools, MultiChannel Framework, Universal Queue, Administration, Broadcast Log	View the log of broadcast events.
Chat	MCF_CHAT_MINLOG_PG	PeopleTools, MultiChannel Framework, Universal Queue, Administration, Chat Log	View the log of a specified chat session.
Event Log	MCF_EVENTLOG_PG	PeopleTools, MultiChannel Framework, Universal Queue, Administration, Event Log	View a log of events and display details.

Viewing Broadcast Logs

The broadcast log page displays detailed information about the any broadcast message that was sent.

Access the Broadcast Log page by searching by message number, queue ID, or REN cluster ID.

Note. To view the broadcast logs, set log_bcast to Yes on the Cluster Tuning page. For this parameter to take effect, click Refresh logging parameters on the Cluster Notify page.

Viewing Chat Logs

Access the Chat page.

If enabled, this log records the contents and events of every chat session.

To enable chat logging, set log_chat_ses to yes on the Cluster Tuning page. For this parameter to take effect, click Refresh logging parameters on the Cluster Notify page.

Details Tab

Select the Details tab.

The Details tab displays detailed information about the selected chat conversation.

Viewing Event Logs

Access the Event Log page.

Domain

The application server domain on which this event occurred.

Time event logged to DBMS

The time that this event was recorded in the database.

Event Type

The type of event, as described in the table that follows.

Task Type

The type of task for this event: chat, email, or generic. CTI events are logged in the CTI event log.

The event log records PeopleSoft MultiChannel Framework events sent to the real-time event notification (REN) server, excluding chat content (which is logged in the chat log). The event log can be used for debugging as well as for system monitoring. For example, you can determine when agents log in and log out, or when the queue server was first notified of newly enqueued events.

Data displayed in the event log depends on the event type. Not all fields apply to every event.

To enable logging of state broadcast events, set logDMPQ to yes on the Cluster Tuning page. For this parameter to take effect, click Refresh logging parameters on the Cluster Notify page.

Time is displayed and searched on in the format MM/DD/YYYY HH:MM:SSA/PM.

The following table lists possible event types:

Name	Translate Table Value	REN MultiChannel Framework Topic	Description
Accepted	ACPD	/agent/<agentID>/accepted Agent 's list of accepted tasks	Agent's list of accepted tasks
Accept	ACPT	/queue/agents/accept	Agent accepts an assigned task
Bcst Admin	BCST	/queue/admin/statedump	Broadcast universal queue information
Contact	CNCT	/queue/contact	Real-time contact (for example, chat)
DB Cntct	DBCT	/queue/dbcontact	Database management system contact (such as email or generic)
Dump Q	DMPQ	/queue/<queueID>/state	Dump queue state information to log
Done	DONE	/queue/agents/dequeue	Done (dequeue)
Forward	FWD	/queue/agents/forward	Forward
Notify	NTFY	/agent/<agentID>/notify	Notify agent of assigned task
Presence	PRES	/queue/agents/presence	Agent's presence change
Restrt Ack	RACK	/queue/agents/restartack	Restart acknowledgement
Read Cfg	READ	/uqsrv/reread/defaults	Reread defaults
Restart	RSRT	/queue/<queueID>/restart	Restart
Unknown	UNKN	UNKNOWN	Unknown REN server event
Unassign	USGN	/agent/<agentID>/unassign	Unassign

The following table lists event logs:

Name	Translate Table Value	REN MultiChannel Framework Topic	Required Argument Value	Meaning
Abandon	ABAN	/chat/<userID>/<chatID>	ps_type=abandon	Abandoned chat session
End	END	/chat/<userID>/<chatID>	ps_type=end	End chat session
Login	LGIN	/queue/agents/loginstate	ps_state=login	Log on
Logout	LGOT	/queue/agents/loginstate	ps_state=logout	Log out
Message	MSG	/chat/<userID>/<chatID>	ps_type=msg	Message
Push URL	PUSH	/chat/<userID>/<chatID>	ps_type=pushurl	Push URL
Timeout	TOUT	/chat/<userID>/<chatID>	ps_type=timeout	Timeout chat session

Viewing PeopleSoft MCF Logs

Diagnostic PSUQSRV and PSMCFLOG traces are written to each application server domain's log directory. The trace level is determined by the LogFence setting in the domain configuration. You can set the LogFence parameter with PSADMIN configuration of the application server domain. The following table lists LogFence setting values:

LogFence Setting	Tracing Level
1	Fatal errors
2	Errors
3	Warnings
4	Level 1 Diagnostic: Logs the queues that the universal queue is servicing, logs new queues that are added, and logs agent logon and logout.
5	Level 2 Diagnostic: Logs most debugging information, except periodic events (such as timer check and heartbeat).
6	Level 3 Diagnostic: Logs everything, including periodic events.

Administering Overflow and Escalated Tasks

To administer overflow and escalated tasks, use the Overflow Administration (MCF_OVERFLOWL_CMP) and Escalation Administration (MCF_ESCAL_CMP) components.

This section discusses how to administer overflow and escalated tasks.

Pages Used to Administer Overflow and Escalated Tasks

Page Name	Object Name	Navigation	Usage
Overflow tasks	MCF_OVERFLOWL_PG	PeopleTools, MultiChannel Framework, Universal Queue, Administration, Overflow Admin (overflow administration)	View a list of overflow events.
Escalation Admin (escalation administration)	MCF_ESCAL_PG	PeopleTools, MultiChannel Framework, Universal Queue, Administration, Escalation Admin (escalation administration)	View a list of escalated events.

Administering Overflow Tasks

Access the Overflow tasks page.

Use this page to manage tasks that could not be assigned to an agent within the specified overflow timeout.

Action Timeout	Displays the time at which the overflow occurred.
Task Identifier	Displays the task identifier.
Task Type	Types include chat, email, generic, and voice (not supported).
Comments	Enter optional text commentary about the task's resolution. For example, note that the customer sent follow-up email that was answered by an agent.
Logical Queue	Select a logical queue to resubmit this task to.
Resubmit	Click to send the task to the specified logical queue to retry assignment. Only persistent tasks can be resubmitted; chat cannot be resubmitted.
Close without Submit	Click to close this task without sending it to retry assignment.
Detail	Click to display additional information about the task.

Administering Escalated Tasks

Access the Escalation Admin page.

Use this page to manage accepted persistent tasks that have been automatically unassigned from agents because they were not closed within the specified escalation timeout. Tasks can be resubmitted or closed.

Action Timeout	Displays when the escalation occurred.
Task Identifier	Displays the task identifier.
Task Type	Types include chat, email, generic, and voice (not currently supported).
Logical Queue	Select a logical queue to submit this task to.
Resubmit	Click to send the task to the selected logical queue to retry assignment. Only persistent tasks can be resubmitted; chat cannot be resubmitted.
Comments	Enter optional text commentary about the task's resolution. For example, note that the customer sent follow-up email that was answered by an agent.
Detail	Click to view additional information about this escalated task.
Close without Submit	Click to close this task without sending it back to retry assignment.

Domain	The application server domain on which this event occurred.
Time event logged to DBMS	The time that this event was recorded in the database.
Event Type	The type of event, as described in the table that follows.
Task Type	The type of task for this event: chat, email, or generic. CTI events are logged in the CTI event log.