Scalability is a commonly used “feature” in all fields of IT and there is no question that it is a real challenge for many IT managers in the near future. IP is a shared best effort technology by default. Like any road – if you double the traffic it will get jammed!It took thirty-five years to get to +8billion IP addresses, Gartner and others predict this will double in less then 3 years. So basically doubling the internet in less then 10% of the time it took to get to where we are today – that will effect your IT systems, but perhaps more importantly it will have massive effect on the everyday IT operations and the need for monitoring.
Lots of devices will be M2M and other “passive” devices not directly involved in your everyday IT. Remember that today we come to work with a laptop, a smart work-phone, a private one and perhaps an Ipad or similar tab, that’s 4 IP addresses – probably 300% more then just two years ago. Soon we will park our car outside the office with an additional one or 2 IP addressees.
All of the device will in one way or another connect to the internet, via your corporate wireless, via the mobile network etc. The point is that bandwidth will be used and this will effect your IP services, bandwidth and security.
But scalability is a lot more then just being able able to handle many devises in your Monitoring system.
I think a core issue is the way your Monitoring solution can be set up to handle the scalability challenge. The way it stores event data and how you manage your distributed monitoring solution.
The traditional approach to just throw more CPU power on the NMS system will not scale and more importantly it will not give you the correct monitoring data to efficiently monitor your network and applications.
A classic definition of TCO is 20% to the initial investment and 80% on maintaining the system over its production years. So ease of use and especially in regards to change management of the system is directly related to the 80%.
Hint: When you look at a new Monitoring system – have a demo on how it scales and how you provision and handle every day changes in your distributed set-up, because if that does not work or looks and feels cumbersome – think again – you will pay dearly in your everyday operation for the next coming years…
Note: the above is really not depending on actual size of your IT – more on your monitoring system architecture.
Scalability and/or performance is the #1 factor in Gartners yearly global enduser customer survey (published january 2012) after the functional capabilities.
My point is that the larger enterprises will hit the “scalability by pure IP numbers” challenge first - however, it will effect all and everyone faster then You might think and plan for today. So whatever size you are today – double or triple the number of IP attached devices in your three year planning. And make sure your monitoring solution does not become a bottleneck in maintaining good overview and control of your applications and network quality.
Read more on how op5 Monitor Scale
Just some thoughts – cheers / Jan
