Amazon and Google made headlines in 2012 for their highly visible outages, proving just how susceptible companies are to IT disruptions, and the negative impact these can have on sales and productivity.
So before jumping into 2013 projects, IT and network managers should establish baselines to understand exactly where they stand. As networks become more complex (virtual vs. physical, wired vs. wireless, etc.), the pressure is on to improve the performance and availability of business critical and customer-facing applications.
IT professionals would benefit immeasurably by setting four key baselines aimed at giving them control of their network:
- Inventory baseline: You can’t control what you don’t know exists
- Performance baseline: Start with the ‘big five’ – CPU, memory, disk and interface utilisation, and ping latency – then gauge key application consumption and optimal thresholds
- Configuration baseline: Understand how current configurations impact security, compliance and overall control of the network
- Bandwidth and data flow baseline: Measure what’s happening on the network, when and how much bandwidth is consumed.
A company’s ability to grow hinges on IT performance and availability, yet many organisations fail to recognise IT’s impact on the bottom line until it is too late. Consider the following:
- The e-commerce application lags, or worse, become unavailable
- Corporate email goes offline and severely impacts on productivity
- Business critical applications such as SalesForce or SAP become unresponsive.
The financial loss could be crippling. No matter how minor or severe, IT disruptions impact everyone. And as IT environments become increasingly complex, the onus falls on IT departments to optimise effectively for performance and availability.
Optimising the infrastructure starts with establishing an IT baseline. This becomes a measuring stick for understanding: (a) how the network, applications and infrastructure perform; (b) where and why performance comes up short; and (c) actionable steps for continuous optimisation. Creating a baseline includes four essential elements: inventory, performance, con?guration and ?ow.
1. Inventory baseline
While most IT managers have visibility into the core infrastructure, awareness of edge devices is much more opaque. Unknown devices complicate network management, as these can consume signi?cant resources and impact the performance of critical IT assets. To get a better handle on the network, there are three key areas to baseline: hardware, systems and applications. Is everything up-to-date and running on the latest rev levels? Have all security patches been deployed? And how does everything in the infrastructure connect?
Understanding the interdependencies on a network is especially important for uncovering and resolving issues quickly. If an employee recon?gures a router by moving it from one sub-net to another and causes a loop in the network, the change can have a catastrophic e?ect across the entire network. Having an inventory baseline makes problem discovery and resolution much easier, and helps to control costs by identifying under-utilised resources that can be redeployed.
2. Establish performance thresholds
Creating performance baselines starts with the big ?ve: CPU, memory, disk and interface utilisation, and ping latency. Network admins need to know how much of the ‘big five’ their mission-critical services and applications consume. More importantly they must know their optimal thresholds. Performance thresholds matter. If a network device has 98 per cent CPU utilisation there is a good chance that device is about to fail, impacting on network availability and performance.
The key to performance baselines is to understand the acceptable threshold levels for each network device and server on the network, and having a real-time alert system for when thresholds are broken.
3. Configuration baselines
Security, compliance and control are on every CIO’s priority list. They are also essential elements to baseline. Looking across devices on the network:
- Are they all running authorised con?gurations?
- Have all security features been enabled?
- Are default passwords still being used?
- Can you generate an audit trail of all con?guration changes?
Misses in these areas could result in damaging security and compliance breakdowns. The most advanced IT departments enforce rigorous con?guration change control policies. They archive authorised con?gurations, receive real-time alerts when con?gurations change, and generate reports answering the critical questions: who, what and when. This makes corrective action – and proving compliance – much easier.
4. Bandwidth and data flow
This baseline helps IT professionals to understand how network capacity and bandwidth are consumed. A complete ?ow baseline breaks down capacity and bandwidth use by users, departments and applications. Optimising network bandwidth and capacity is critical for enhancing performance and productivity. IT managers must understand what’s happening on the network, and how much bandwidth is being consumed. The end goal is to ensure that business- critical applications have the bandwidth they need to operate at maximum e?ciency.
Understanding how much network capacity employees’ use also impacts the budget. From a bird’s-eye view, the company might seem to need more bandwidth, while in reality it might save 30 per cent of existing bandwidth by identifying unauthorised use of bandwidth hogs like YouTube.
Today’s IT environments are dynamic and complex. Changes occur every day that affect performance and availability. But a baseline of assets and performance thresholds gives IT a measuring stick they can leverage in real time to enhance overall network performance and efficiency.