List: Session at VeeamOn 2015 looked at how to improve efficiency in a virtualised environment.
From Las Vegas, Nevada, software company Veeam hosted a session to explain how businesses can do virtualisation monitoring in Microsoft’s System Center.
The presentation was headed by Clint Wyckoff, Veeam Microsoft Evangelist, Dave Kawula, Microsoft Most Valuable Professional (MVP), and Michael Stafford Veeam Solutions Architect, Veeam.
CBR lists the main points on how to monitor in a virtualised environment.
1. Do something to join the dots
In the first place, system centre operators need to map and track relationships and dependencies to get the right insights into and from the system.
The key elements on this stage are availability and status of the applications, VMs, hosts, storage, network and hardware.
Wyckoff said: "IT exists for one sole purpose: to serve the business and the business is applications. Virtualisation really changes everything because it combines everything into one thing. There are a lot of things that can go wrong and you need to keep an eye on."
2. Visualise key metrics
The next top tip is for operators to visualise as they virtualise. Operators need to be able to have at their fingertips a series of dashboards and reports to better understand what is happening with the applications. "Context is key," said Kawula.
The hosts said that monitoring host CPU counters is inaccurate. Memory, analysis and storage monitoring are three pillars that will help improve efficiency in a virtualised environment.
Stafford said: "It is all about reducing waste. Waste increases your costs. [Storage latency is] Definitely something folks have a problem with in a VM environment."
3. Learn to share
Sharing between partners is a critical aspect the industry is currently starting to look into.
A key metric in this area is operators’ over-commitment which sometimes becomes a barrier to sharing.
Kawula said: "Everybody has to share, understand resource sharing and over-commitment. If I have a VM it is important to have a holistic approach to it because we are sharing things."
4. Get a crystal ball
Forecasting what will happen to the system will prevent the system from failing in the first place allowing for it to stay always on and reduce any associated costs with downtime.
Wyckoff said: "Look to the future, use forecasting trend monitoring."
5. Knowledge is power
When combining all the above, System Center operators will have at their disposal a new stream of information to help them stop guess work.
With monitoring tools in place, knowledge about the whole ecosystem will help empowering those responsible for the system to better understand when things work and when they fail (or are about to fail).
Wyckoff said: "Any time you have an alert, there is not guess work involved."