The Problem

Many of us operate computer and peripheral networks of increasing size and scope.  Unfortunately, many DSL and cable-based Internet connections occasionally fail in unexpected ways.  The convenience of modern wireless networks also comes with a hidden cost of complexity.  Sadly, most users don't have the ability to diagnose the location of a failure.  The goal of PingView is to allow both experienced and novice computer network users to monitor and troubleshoot networks.

PingView does real-time monitoring of small- to medium-sized wired and wireless networks using a variety of testing methods such as the well-known "ping" and selective TCP port probing.   (Most modern routers, modems and wireless equipment respond to standard "pings".)  Instead of specifying network elements in a flat list, PingView allows you to indicate the network "topology", or the relationships between elements.  For example, your router should respond to a "ping" before your DNS server is tested, since a router failure will block all further access to the Internet.

When a failure occurs, PingView clearly marks the location of the failure.  The target information can contain the location of the device and the corrective action to be taken (restart, call someone, etc).

I use PingView in my own home to support both "sides" of the network: the "outside" connection through my modem and DNS out to the Internet, and the "inside" to monitor the wired and wirelss legs of my own network extensions.

Getting Started: the Wizard

PingView has a built-in "wizard" that attempts to discover the basic configuration of your network.  After developing the basic configuration with the wizard, you can use simple "drag-and-drop" operations to indicate which devices are dependent on other devices.  In PingView, each testable unit is called a "target"; an ensemble of targets is called a "collection". The image below shows the result of the wizard in a simple network. 

The collection of targets is displayed on the left (collection) pane and detailed status information is in the right-hand pane.   The right-hand panes show (from top to bottom):

 

The Wizard lists network elements that are not part of the originating computer's path to the Internet last (eg., 192.168.1.72).  The originating computer's gateway is listed first.  Its dependent devices are its DNS server and one actual Internet host for testing (the National Timekeeping Service is used by default).   When this target collection is run in PingView, each entry will be "pinged" or tested to see if it is "alive".  The failure level will be indicated by a single red line, indicating where in the data path the ping failed to receive an answer.

Using the wizard, targets in the local network are usually shown as IP addresses unless the network (machine) name is available.  You can use IP addresses or domain names for Internet targets or even for "intranet" targets if your network has its own DNS server.

If the originating computer is connected wirelessly (IEEE 802.11x), the computer's wireless "access point" will appear in front of the its gateway (router).

At this point you may run your collection; see Running the Collection for more details.

What PingView Does

PingView probes your collection of targets to see if the network should be considered "up".  If every target marked as "vital" responds in a reasonable amount of time, the network is considered operational.  If not, the display is changed to indicate which target or targets failed.  After a pre-defined number of failures, the network is considered "down", and sounds may be played (or scripts run) to notify network administrators.

If a target fails to respond, its dependent (lower level) targets are not tested.  This means that not only does PingView report failures more promptly, but its error indications are clear because only the first failure in a causal chain is reported.

Since PingView is designed for smaller networks, each target should contain information about where the device is located and the proper method for restarting or correcting problems.   This allows even novice users to go directly to the source of the problem and troubleshoot the network correctly.

Reporting Failure: "Vital" and "Ignore"

PingView segregates targets into two types: vital (displayed with a green icon) and non-vital (displayed with a grey icon).  Every vital target must respond to a "ping" during a testing cycle or the network is considered "down".  After a pre-configured number of "down" cycles, the network's outage is repored using message, sounds and optional script execution.

Devices that are occasionally or frequently off should not be marked as "Vital".  If they are not marked as "Ignore", they will be tested, but their failures will not cause the overall network to be reported as "down".   In addition, even if a target it marked as "Ignore", you can still right-click on it and choose "Ping Now" from its menu.  In this way you could, for example, check for rarely used web or FTP sites before you actually try to use them.

PingView and Dynamic IP Addressing

PingView is primarily intended for network components that use fixed IP addressing.  However, most small networks and many large ones rely on DHCP (Dynamic Host Configuration Protocol) to assign IP addresses to computers when they start.

In a Windows network, each machine is given a machine name, also known as a NetBIOS name.  PingView records this information if it's available.  When PingView loads a collection, any targets that have machine names are treated specially.  For each of them, a query is made to the network software to return the current IP address for each named machine.  In this way, PingView can easily operate with dynamically addressed devices.

However, machines using DHCP may sometimes be restarted while PingView is running.  Rarely, this will result in a machine getting a different IP address than it previously had, hence confusing PingView, which will report the old IP address as unreachable. To correct this problem, stop PingView cycling and choose "Update Network Configuration from the "Actions" menu.  This will briefly poll named machines to find out their new IP addresses.  After that you can start PingView's cycling again.

It should be emphasized that the primary purpose of PingView is to determine the status of networking components and Internet sites, not to query the status of machines that may be on or off at any given time.  Overspecification of the network will result in many "false positive" error signals from PingView.   In other words, limit the devices (targets) known to PingView to components that everyone uses.

Getting Ping Responses

Many firewalls and Internet security software packages automatically configure computers to ignore "ping" requests; this is because pings can be used by hackers to locate and target your equipment.  When you use PingView, however, you have to make a decision about whether the risk of being hacked is outweighed by the need to monitor and diagnose your network.  Most network administrators are of the opinion that answering pings is quite safe.

To get a computer in your network to respond to pings, you must bring up the configuration program for its active firewall or Internet security software.  Then locate the place where you can create or enable network traffic rules.  Standard "pings" are usually referred to as "incoming ICMP requests", or words to that effect.   Most firewalls have a single, simple rule that you can enable (if it's present but disabled) or create to allow the machine to reply to pings.

If you find that your firewall or security software makes this difficult, I recommend you search Google with a query like "enable ping response using <name of software package or firewall> on <name of operating system>".  

Types of Ping Targets

As stated above, the PingView application attempts to "ping" each target in the open collection.  There are different types of "pings", however.

Each type of pingable target allows you to configure an address, a displayable name, a device location, its repair operation and whether it is vital (considered essential) and whether it is to be ignored.   You can also configure special "management" URLs that are used when launching a web browser to control the targeted device or system.

ICMP Pings

This type of target is simply a fixed IP address or domain name that is pinged by sending a standard ICMP echo request message.  The IP address for a domain name is usually fixed but can vary in a DHCP-based network.  PingView always uses the address field ("hostname") as given with pinging, so that dynamic DNS servers can handle changing IP addresses.

ICMP Pings to Machine Names (via NetBIOS)

Machines in many SOHO (small home, small office) networks use dynamic host addressing, known as DHCP.  Under DHCP there is no guarantee that the device will retain the same IP address over time.   Everytime the device or system is restarted it must again request a unique IP address from the DHCP server, which is usually in the router.

Microsoft systems have a networking technology known as NetBIOS.  This system runs on top of the usual Ethernet protocol and creates a mechanism for addressing systems and devices by a user-readable name.  All Microsoft Windows systems support this, and many products from other vendors such as printers do also.  You can configure an ICMP ping target to utilize the NetBIOS or machine name instead of an IP address.  In this scenario, PingView relys on the Windows networking system to convert a machine name to an IP address.

TCP Pings to Particular Ports

PingView can also be used to determine if certain service ports are open and available on systems.  Most web servers, for example, use port 80 for all traffic.  You can create a target in PingView and mark it as a TCP target, specifying the port to be tested.

In this scenario, PingView will perform a TCP "open" operation on the given system and port; if it succeeds, the connection is immediately closed and no data is transferred.   If this succeeds, the service port is considered to be operational.

PingView Servers

As documented elsewhere, a PingView collection in one location can request a copy of the collection running in a remote machine.  This allows PingView to report the status of systems located behind firewalls and routers that would otherwise be inaccessible from the remote location.  It also allows PingView instances to receive status information from remote locations.

Group Targets

There is no limit to the size of a PingView collection; in principle, a collection could contain hundreds of targets.  In order to simplfy reporting and diagnosis, you can create special targets known as "group" targets.  These are really surrogates and do not contain any real addressing information; they are used to organize the PingView display into working sets to simpify monitoring and diagnosis.