Go to Chapter 2    :  Chapter 2.2        Contents

Section 2.1: Understanding the Graphic Elements

2.1.1 DS0 Channel Graphic Elements

There's quite a lot of information to pack into the small real estate of the GW_Monitor window. The Graphics has been designed to show as much as possible at a glance. Each DS0 channel will have three elements:

The larger square is the Dial Peer Square, it will show the Dial Peer color. There's an algorithm that will assign automatically a different color to every Dial Peer ID. As soon as the a match is found the call is attempted and the gateway reports it to the GW_Relay and this last one to the GW_Monitor that will light the square with the corresponding color. All channels featuring the same color at a glance can be seen to be calling the same peer. This is the default color assignment, but there's another, using the Menu item View/GroupColors/For/[Peer ID or DS0 groups], you may choose to assign colors according to the DS0 instead of the default Peer ID's. (See details in Section 2.3.8

On the center of the square you may find a circle, this would be the Sound Quality Circle, it indicates poor sound quality. If the Latency and Gap values go beyond the specified limit values, the circle will start showing, the greater the circle the poorer the quality. The color of the circle will go from blue to red. Meanings are show below:

The diameter of the circle is proportional to these two effects. The contribution of each is going to be expressed as a fraction of the size of the of the Call State Square. In other words, in the worst case, the diameter of the circle (in magenta) can grow to be twice that of the Call State Square side. The contribution of Gap will go from 0 to 100% only when its ratio to the one specified in  PLAY/GAP_RATIO_MIN does it from 1 to 11 . The contribution from the Latency will be grow from 0 to100% when the RTT  goes from 1 to 4  times the value specified in RTT_MAX .

Note: When the silence suppression feature is activated in a gateway, the CODEC will insert silences after a command instead of playing back silent packets. This saves considerable bandwidth, but the gap is no longer related to sound quality but to the nature of the feature. Calls using Silence Suppression may show at the beginning big circles, fortunately in these cases, gap can be greater than playtime when a party is listening, indicating to the GW_Relay that the features is being used. In this case, total packet loss to received packet count ratio will be used instead. 

The Call State Square can be found either on the lower right cornet of the Dial Peer Square or on the upper left corner: Position indicates call origination, PSTN or IP respectively.

The color of the Call State Square indicates the above the color code. Below a description of the states

Idle                No current connection but ready

Setup            Call has been setup but still no response from the destination.

Signaling        Call is either being dialed or is alerting

Connected    Charge time has started, parties are talking

Blocked        Channel has been et out of service, span of synch, span is down, span inexistent

Busied-out    The channel has been busied out, either manually by an operator or automatically by the gateway. Data channels in PRI spans also show in blue

When you hover your mouse pointer over a Call State Square, the Border of the Dial Peer Square will lit in white if the Dial Peer is dark or turn black if the color is light. This will indicate that the channel is in focus and the information in the Call Info Text-box (See below) and in the Tool Tip Text will be referring  to it. The Tool Tip Text shows the Channel Number, if the it working a inbound (R for receiving) or Outbound (O) and its Dial Peer ID preceded by VoIP or Pots in case of inbound or outbound respectively. Channel numbers are assign according to the following formula:

Channel Number = (100*(Span Number (1 to 4)) + Timeslot Number (1 to 24)

Inside the Call State Square, you may find the Mark (X). A Mark is not information from the gateway, it is you who marks a channel to select it to do an operation on it. To toggle the  mark on a channel, you just click on its Call State Square. You may also toggle the mark on a group of channels by clicking and dragging a Marking Rectangle (see below).

Possible operations are requesting traces for debugging or taking some action by using the Busyout feature

Occasionally you may find a "?" inside a  Call State Square, here's how and when that happens. While a inbound call is being setup into a DS0 channel, but still no routing information is available, the Call State Square will show the "?". It will disappear as soon as it gets routed, but what if, for what ever reason it can't be routed? Then, it will remain there, even after the channel is back to idle, warning that something went wrong. The "?" indicates an IP link failure, a dial peer mismatch, a remote gateway not responding, etc.

2.1.2 Text Boxes and Lists

The Call Info Text-Box shows the information about the call channel the mouse is hovering or, if the mouse is not over any particular channel,  the latest call information arrived. You may select what information of all the available is displayed with parameter CALLINFOMASK A Call Detail Window is available that show all the Call Info Text-Boxes for a span. In a PSTN to PSTN call, the IP field is replaced by the channel number of the other call leg. If the call is associated with an account, as for instance in a prepaid IVR application, the account number with show after the IP field in the form "Acc = xxx".

The Last Alarm Text-Box show the last alarm to arrive and the local computer time (not Relay time, not gateway time) at which it did. Alarms before that can be found on the Alarm List in the Alarm Tag. The alarm messages can be copied to the clipboard an pasted in any other document, you can select and delete and you may specify the maximum number of alarm messages to keep on the window. For more information go to Big Bro Watchdog Alarm System

2.1.3 Gateway and Interface Information

Starting from the right, is the "Uptime" of the gateway, is shows the elapsed time from the last reset in years (yy), weeks (ww), days (dd), hours (hh) and minutes (mm).  Uptime is not to be confused with the Gateway Time, that, you can't miss it, is a big text box labeled GW Time this is the reading of the gateway's internal real-time clock..

Next to the left comes the name of the main interface as set in the GW0Relay configuration file with parameter INTERFACE_NAME  The label below show the bandwidth available to that interface as set by parameter INTERFACEBW in the same file. 

Following to the left, there are 3 Bar Indicators. Each indicator has two bars, the top one shows average and the bottom, the last value. 

The first of the bar indicators to the right is for the CRC (Cyclic Redundant Check) error count. In absence of errors the indicator shows no bars at all. If errors were detected, the upper one will jump to 50% of the total width while the bottom will show the ratio of the last count to that of the maximum detected. If it goes to 100% (full bar) it would mean that the problem is still as bad or getting worst, if the bar gets smaller then it could be coming back to normal.

The next two Bar Indicators show the In and Out Bandwidth Usage. The top bar is the average bandwidth usage since the GW_Monitor was opened. This is the ratio of the received or transmitted bits to the maximum bit count that could have been received or transmitted in that time according to available bandwidth. The bottom bar is the current bandwidth usage, the size of the bar show the ratio between the actual bit count to the maximum possible bit count in that same interval.

If you double click on any of the Bar Indicators a graphic with the BW Usage History of the displayed variable will drop down. If you move the mouse over the graphic, the tool tip will show the time at which the pointed count was reported. The graphic has a total of 50 thin bars, it evolves in the following way: the first 50 samples are mapped into each of the mentioned bars; on the fifty first sample, all even bars will be copied into the first 25; thin bar 26th will be set to this mentioned 51st sample; the rest 24 bars will be set to 0 length; the next sample will be skipped and the next will go into the bar 27 and so, every two samples until the 49th bar; now the same as the first cycle, all even bars to the first 25 and now every 4 samples .then 8, 16..and so on. The result is that the graphic always covers the whole history from the opening time of the GW_Monitor, no matter for how long it has been opened.  

There are labels and check boxes to the right, one of the labels should be in read indicating the Master Clock, the rest of the span are locking to the clock of this board. The check box is for debugging purposes, see Section 2.4.3 Tracing. 

2.1.4 GW_Monitor Protocol Status

To the left of the BW Usage Indicators, there is a Watchdog Indicator that shows at a glance that the system is up. The bar length is reset to 0 by the first GW Time update once it goes beyond half length (actual time is determined by WATCHDOG_TO parameter in the ini file). If the gateway crashes or its link goes down the clock update won't come and the Watchdog Indicator will go all the way to the right end. Upon getting there it will turn red and start to blink. An alarm message "Gateway not responding" will be issued, the Sound Alarm activated and Watchdog Pages issued. The GW_Monitor sends a  "are you OK?" message every 30 seconds, if a reply to such message has not been received either, then the absence of clock update is not due to a gateway crash nor a link down, but to a GW_Relay problem or to its link. In this case the Alarm Message would be "Gateway Relay not responding". 

Next to the left is the Pending Message Counter, messages coming into the socket are parsed and saved into a buffer, pending messages are processed in the order they are received. This buffer holds up to 1000 messaged. Normal operation should show a 0 that may occasionally blink into some small number and then back to zero.  I you see a persistent Pending Messages Count, then the  window you are looking at, is not up to date with information from the GW_Relay, if you see a number greater than 1000, then you have already lost information. GW_Monitor has been observed to lag behind in the following situations:

Last to the left is the Acknowledge Indicator, when you request information from or send action commands to the GW_Relay, it responds with an Acknowledge Message when successful or a not acknowledge if it fails. The GW_Monitor protocol has one character commands followed by parameters. The Acknowledge Message consists of this command character preceded by a (+) or a (-) to indicate success and failure respectively. The Acknowledge Indicator will show the mentioned command character and will blink in green when (+) or in red if (-). This last message will stay there, with its normal gray background, until the next command is acknowledged.

Go to Chapter 2    :  Chapter 2.2        Contents