"It's the biggest miss we see," says Tom Elowson, president of virtualization cloud provider Acxess. "We have the bandwidth conversation with potential clients every day. If they haven't analyzed their existing usage and started to calculate the potential impact, we usually push back."
Start with the outbound volume to reach the resource, and take into account back-end traffic to update data. Bandwidth calculations also need to factor in data and user growth over a five-year period, same as ROI calculations. Get solid trending stats on usage and volume over the course of several weeks. If you don't, you could be looking at a major fumble.
Monitoring: Watch And Learn
Thirty-nine percent of poll respondents say they don't monitor their cloud vendors, while an additional 40 percent rely on basic "up/down" tools that are no better than a periodic ping. The latter group's sole advantage is they'll have a 30-second warning before the complaints start rolling in.
How to stay on track? First, invest in data flow monitoring internally. Less than 15 percent of respondents have systems in place that monitor application and transactional throughput. Basic status alerting is nice, but you need to be watching your network data flows and have established performance levels for every application before you add an external cloud.
Once your house is in order, connect with your bandwidth provider and establish ground rules around monitoring of traffic, your lines and how you share data. Set up remote monitoring points outside of your main office. Assemble a set of cloud-based monitoring tools. Yes, a cloud app to watch your cloud apps. Go beyond the basic utilities that Amazon, GoGrid, Google and others provide to add overall monitoring of all Internet traffic.
