100% WAN connectivity

[rauldpaglinawan]

How do I ensure that I will have a 100% connectivity across the Wide Area Networks. What are to tools that we need to execute?

[wong34539]

I would always recommend nothing more than a simple ping and Extended ping command for ensuring this connectivity.

From your location, you can use ping or extended ping to connect to a server/ a pc in the other end of the WAN environment.

Use Trace route command to trace the path to a destination.

[Patrick Laidlaw]

Hello,

If your looking for testing tools to help determine uptime there are quite a few of them. Whats Up Gold is a fairly common commercial software tool which will ping each ip address to see if it is up and responding it has several different ways to notify you.

A GNU tool which does much more is Nagios.

If your looking for a way to keep your uptime at 100% you'll have to look into redudant connections using different carriers and additonal hardware.

Patrick

[twojciac]

The short answer is that there are many tools which offer varying levels of monitoring and paging integration. CiscoWorks, HP Openview, and others mentioned here are offer snmp monitoring services.

The long answer, 100% is a nice goal. High availability requires process, planning, and metrics equally. Here's a whitepaper on some of the details of how my team achieved 99.999% LAN availability. The heart of it applies to WAN, which with proper design you can achieve 99.99% availability.

http://www.cisco.com/application/pdf/en/us/guest/tech/tk769/c1550/cdccont_0900aecd800b29ac.pdf

[mheusinger]

Hello,

there are several possible answers to this. The first question would however be: how do YOU define "100% WAN connectivity"?

You can not rule out power outages, earth quakes and other things that could bring down your service. The most dominant reason for downtime is however maintainance.

So if you include all this, there is no 100% availability from a mathematical point of view.

Now you can exclude some things. In fact you could exclude everything bringing down the service and have 100% availability BY definition!

So how does your availability definition look like?

To find out which uptime a WAN has, some people use Cisco Service Assurance Agent or SNMP polling and the like. Again this gives no absolute accuracy. In case you poll every 5 minutes you will not likely detect an outage lasting 20 seconds.

Now polling every second in the WAN is not an option usually, because you generate simply too much traffic.

So again you need to think about the granularity at which you like to see availability. This usually comes with looking at application needs.

Hope this helps. Please rate all posts.

Regards, Martin