The top status panel shows a quick status report for the current situation.

The first line shows information for all selected queues as a sum, while if there is relevant information for a specified queue it is displayed in a separate line. If an alarm is triggered for one of the numeric values displayed, the relevant cell turns either yellow or red.

The displayed fields have the following meanings:

Please note that, as agents are not linked to a specific queue save for the moment they are actually talking to a caller on the queue, the agent information is computed for all agents on the Asterisk server and not for specific queues, unless the "Members" option in the "Agents" dropdown is selected.

A list of calls flowing through the selected queues is presented on the middle table. If no call is present the table is displayed empty.

When a call is processed, the following fields are shown: - Queue: the queue that is handling the call; - Caller: The Caller-ID, if available; - Entered: The date and time the call entered the queue system.

If the call is not answered yet, the "Waiting" field is displayed in red and is calculated according to the current date and time of the server. Depending on what type of information is present in the database (ATTEMPTS or RINGNOANSWER), is possibile to have information about the last agent not picking up the call or the actual ringing agent’s phone. The default configuration works with a standard Asterisk configuration and lets able to have RINGNOANSWER information. Please refer to Configuring Asterisk for QueueMetrics for the how to configure Asterisk generate ATTEMPTS information.

When a call is answered, the "Waiting" field tells the time that the caller had to answer; the "Agent" field shows the agent (or terminal) the caller is talking to and the "Duration", in red, is the current call duration.

If the call is ongoing and connected to an agent, moving the mouse on the wizard icon at the end of the line, some icons, like Call Monitor and VNC Monitor icons may be present. By clicking on one of these icons you activate the specified monitoring (see below).

If the call is ongoing and you have the special grants to do so, a red scissor icon might appear, moving the mouse on the wizard icon, to allow for brute-force call closure. See the section Closing ongoing calls for further details.

As soon as a call is completed or hung up, it exits the Calls panel.

This panel can be turned on or off through the "Calls" dropdown on the top of the page.

The "MOH" field shows the amount of time a customer is on hold with music, during a call. If multiple HOLD instances took place during the call, this filed will show the total "on hold" time.

The "Srv" column is used only in cluster-based environments to tell you on which server the call is being handled.

The last field contains a "wand" that on mouse-over displays a drop-down menu which allows to perform the following actions:

To have precise realtime reporting a perfect clock synchronization is required. If your QM is on a different server, make sure the clocks are exactly aligned or you may see strange values in all fields. The NTP protocol offers excellent clock synchronization precision and is available on most operating systems.

A list of available agents for all queues is displayed in this field. For each agent, the name, last log on and extension, if logged in via call-back, is provided.

A graphical indication of the status of each agent is shown using a coloured dot, where the following cases are possible:

This panel can be turned on or off through the "Agents" dropdown on the top of the page.

The Queue(s) field shows the queues an agent is logged on to. This is meaningful only for agents who log-in on a queue-by-queue basis using the AddMember command in Asterisk. If an agent logs on to all queues he’s enabled to work on, a small database logo may be shown, telling the viewer that the agent is linked to queues through the Asterisk’s configuration.

As the queue_log file usually contains no information on what queue(s) an agent is a member of, usually all agents are shown when they log on, no matter to what queue they will work on. This might be a problem for larger call centres, so it is possible to see only calls and log-ons of agents that are a member of the current queue. Membership is set by clicking on the "Agents" button of the queue settings page. Make sure your queue membership data is up-to-date before clicking this button!

If the "All" option was selected in the "Agents" dropdown , all agents logged in on Asterisk will be shown, no matter to which queue they belong. When "Members" is selected, instead, only agents defined for the selected queue will be displayed.

By turning on the property "realtime.show_incorrect_queue_sets", the Queues field will display each queue with a different color code:

The On Pause field will contain the time the agent went on pause; if the agent is using a pause code to mark the reason for going on pause, the decoded pause code is shown as well. A switch can be used to show the pause start time either as an absolute hour or a time increment.

The Srv field will contain the server identifier for QueueMetrics running in cluster mode.

The Free Since field will contain the length of period since the last taken call.

The Last call and On queue fields show the start time of the last call the agent handled (which is latest) and on which queue the last call was. This can be useful to diagnose queue strategy problems that lead to unfair call distribution, or agents having problems with their telephones and therefore not taking calls correctly.

The IVR field shows the call time spent on the IVR process before the call enter in the queue.

Waiting time and Duration fields are related to the last call taken.

The "wand" icon mouse-over displays a drop-down menu which allows to perform the following actions:

The last icon is a direct link to reports. Clicking on this icon will run an historical report by populating the query with parameters similar to what generated the realtime panel results.

Since QueueMetrics 16.10.17, a new optional field is available for the Agents Currently logged in table. This field is called Paused For and it shows the elapsed time of ongoing agent’s pauses. To enable this field one needs to enable the realtime.relativePauseTimes System Preference. See Appendix C: System preferences.

Since QueueMetrics 18.04.3 a new field is available, named "Since". This field is located right next to agent status colored icon. The number of seconds listed in this field represents the time that has elapsed since the agent entered it’s current state.

This block reports realtime figures related to each selected queue. All values are aggregated by queue except for the first row where calls for the whole selected queue set is reported. On each column a user could find:

By clicking on the last icon, a user can start an historical report with query parameters matching the realtime page parameters that generated the selected row.

This block reports realtime figures related to tags associated to processed calls. All values are aggregated by tag except for the first row, where calls for the whole tag set is reported, and the second row, where only calls with no tag are counted. On each column a user could find:

By clicking on the last icon, a user can start an historical report with query parameters matching the realtime page parameters that generated the selected row.

This block reports realtime figures related agent sessions and calls. All values are aggregated agent except for the first row, where lost calls are shown. On each column a user could find:

By clicking on the last icon, a user can start an historical report with query parameters matching the realtime page parameters that generated the selected row.

You can also assign each agent to al Location, i.e. a group of agents working together that you want to monitor as an unit. They might be a supervisor’s team, or people working in the same building, or maybe in the same location for big multinational call-centres. This way you can avoid being cluttered with information about all agents working on the selected queue(s)and only see those you are actually interested in.

Note that when you filter by agents or locations you may see calls being queued and then disappear. This happens because all calls that wait to be answered on a queue are shown, but answered calls are shown only if the answering agents is a member of your defined filter conditions.

Imagine you have two groups of agents, one working in NY and the other one in LA. You are the supervisor of the NY group, so you are filtering by location. You see a call entering your queue and then disappear. The reason why might be that it was answered by someone in LA, so QueueMetrics filters it out for you.

It is also possible to have QM ask for a location to be selected in any case for Real-time reports. As locations can be key-protected, this is a handy way to determine which agents can be seen by which supervisors. In order to enable this feature, you must set the realtime.assignedLocationsOnly property. When this is set:

It is possible for you to set up both Unattended Call Monitoring and VNC Monitoring.

Unattended Call Monitoring makes it possible to listen to an ongoing call from an agent; by clicking on the small telephone-shaped icon, that will appear moving the mouse under the wizard icon, a popup will be shown where you may enter your current extension or PSTN telephone number.

As soon as you confirm the entered data, your telephone will ring and you will start listening to the ongoing call between the selected agent an the caller.

Depending on the kind of PBX platform you run, you may have one or more of the following options:

In order to set up this behaviour, please see the section Listening to live calls: Unattended Call Monitoring. It is possible to use different PMs to handle different audio needs - see Listening to calls using Pluggable Modules (PM).

It is also possible to set up the system in order to allow the real-time monitoring of the agent’s screen using VNC. If this feature is enabled, a small screen will appear close to the agent’s name; by clicking on it, your selected VNC application will be launched and you will be monitoring the agent’s screen.

In order to set up this behaviour, please see the section Enabling VNC Monitoring of this manual.

Running a large set of real-time pages / agent pages / wallboards may impose a significant burden on a QueueMetrics system. But sometimes you have no choice. The best way to run them is to:

In order to further reduce the load above, since QM 23.09 there is a new unified architecture (codenamed Presentialist) that aims to keep an up-to-date picture of the whole system and just send a snapshot to all interested parties. So, for example, an agent page and a real-time page sending in a status request at the same time would share the same underlying analysis.

This usually works seamlessly, but there is an important limitation - as all calls are processed no matter on which queue they appear to be on, if the same call is erroneously tracked multiple times (e.g. by a queue and by the Uniloader outbound tracker), the resulting logs happen to be unreadable. In order to bypass this issue, a new property realtime.ignoredQueues was introduced - it lets you blacklist a set of queues for display on real-time/agent-page/wallboard.

On older systems, you may also want to make sure that you are using the following defaults:

The real-time live page can be accessed by clicking on the "Live" tab next to the "Realtime" tab.

This page is not built from the queue_log data as all other information reported by QueueMetrics, but it' s read right from each Asterisk box’s Manager interface. So what you see in this page is the status of each Asterisk box, as reported by itself.

This feature is still less developed than the Realtime page, but still can be pretty useful.

On the top panel, for each queue on each server, you will see the following pieces of information:

In the calls panel you see the following pieces of information, sorted from oldest to newest by call start-time:

The Wait and Talk times cannot be distinguished at the moment.

In the agents panel you can see the following pieces of information:

The last panel details the status of each server making up the cluster. If a server is not correctly set up, it will appear as KO.