Monitoring the Internet of Everything
Fresh off Cisco Live 2014 in San Francisco, the Internet of Everything (IoE) is top of mind. Why is this topic so important (other than because Cisco tells us it is)?
Simply put, the IoE will have a significant impact on your ability to monitor the performance of your network and data center. This is especially true in the case of Service Providers (home monitoring systems), Healthcare (patient monitoring devices), Manufacturing (automation and sensors), Finance (handheld card readers and POS systems), Facilities and Building Management (climate controls and elevator monitoring), and perhaps even Automotive (on board diagnostics).
There are five main impacts you need to concern yourself with:
- Massive influx of data – At its heart, the IoE is an issue of scale. Cisco predicts 50 billion connected devices by the year 2020. This signifies a massive wave of device-generated, machine-to-machine data you’ll need to monitor. It doesn’t matter if you’re a service provider, large enterprise, or small- to mid-sized organization. Scale is relative. The volume, variety, and velocity of data traversing your network will increase by many factors. And what about transaction data, climate and energy metrics, footfall, or other time series data? When overlaid with your traditional infrastructure performance metrics, these data sources provide new layers of insight into current and future capacity demands. Are you capable of ingesting, alerting, and reporting on massive volumes of disparate data at speed?
- New devices coming online – Devices today are smarter and more connected. And they’re coming online at an accelerated pace. These new IP enabled devices – such as climate and energy controls – will penetrate your data center and business facility. How long will it take you to get SNMP certifications for monitoring new devices as they come online in your environment? Or if SNMP is not an option, will you be able to get at metrics from new protocols?
- “Bursty” traffic – Evidence supports that today’s data traffic has become much more “bursty” than years past. Polling your environment at five minute intervals is akin to falling asleep for the better part of a movie and then trying to figure out what happened while watching the end credits. Even one minute polling cycles can mask spikes in traffic that challenge your current capacity. Can you achieve sub-minute polling of your network – at 30 seconds or less – to better understand the nature of your traffic and future capacity needs?
- Heavier reliance upon the Cloud – The IoE coincides with another common trend… the march to the Cloud. As your organization migrates applications and services to the cloud, you lose control of aspects of those applications and services. Yet you remain responsible for monitoring their performance and ensuring delivery. How do you monitor performance of something you no longer own in its entirety?
- Rapid adoption of IPv6 – The IoE is perhaps the greatest driver of IPv6 adoption we’ve seen (well, in addition to the US Federal Government). IoE is not an IPv4 world. The proliferation of connected devices demands expanded network addressing. Yet, many performance monitoring platforms remain incapable of monitoring native IPv6 environments. In dual stack environments, can you monitor both IPv4 and IPv6 data within a single NetFlow report?
If you’re not contemplating the impacts of IoE today, you risk visibility gaps over your infrastructure. Visibility gaps mean you lose the ability to monitor and respond to performance and availability issues within your network and data center. Visibility gaps lessen your ability to enforce strict SLAs and they put your customer and end user experience at risk. Not to mention revenue and productivity.
Is your performance monitoring platform well-equipped to handle the inevitable impacts of the IoE?