Any telecommunications service uses a huge amount of infrastructure. Too often, the infrastructure is monitored by silos of service, without being able to look across the company to understand what needs to be done to provide the highest quality of service with the greatest uptime reliability.
Network monitoring along silos of business simply doesn’t work in a 24/7 service like telecommunications. Instead, unified monitoring helps maintain the highest quality of service. This approach controls expenses and keeps teams from needlessly pursuing events that have no effect on reliability.
Some forward-thinking telecom companies use unified monitoring already. In New Zealand, 2degrees uses enterprise monitoring for all of its business systems. It also provides its network operations center with incident detection and resolution of issues that could otherwise harm service delivery, infrastructure, networks, and applications.
According to Klaus Ahlers, IT OSS Manager for 2Degrees, “Unified monitoring allows us to improve our customer service by understanding and monitoring tractions across a variety of systems.”
With unified monitoring in a 24/7 service like telecommunications, there are five things to keep in mind:
- Customizable threshold policies
- Proactive notifications
- Real-time service dependency modeling
- Root cause analysis
- Open support for existing devices
Let’s take a closer look at each.
Customizable detailed threshold policies
How can your IT organization separate important events from “noise”? The most important thing is context – the relationships between resources, and their dependencies. With that understanding, you can accurately identify the threats to delivery service and those you can safely ignore.
To better identify and resolve service issues, you need a real-time view of system events, fault notifications and status changes, across you entire infrastructure. Give up monitoring by silos of servers, networks and storage. That gets you only very limited insight into how events may lead to degraded or disrupted service.
For effective collaboration, you need to share insights into IT events with other team members. Event, performance and configuration data have to have triggers that automatically notify the appropriate team (via email, text or other messaging) about issues that can affect service.
For faster, more accurate IT operations, your systems must be integrated with provisioning systems, orchestration systems, service desk applications, and configuration management databases. Look for solutions where events can trigger monitoring system commands, or run diagnostics automatically. For complex issues, you need to be able to activate external orchestration systems to make sure you are able to recover from the event.
Real-time service dependency modeling
IT performance problems can hurt revenue, so it’s important to discover possible service problems before your customers do. With customer-facing application services, you need near real-time awareness of disruptions, continuous service quality monitoring, on-demand configuration modeling, dynamic performance and event monitoring.
It sounds basic, but your data center teams need to understand which infrastructure elements support which IT services. Monitoring requires modeling of infrastructure components based on their relationships to services. Administrators can tell when there may be even a potential risk to service. That not only reduces downtime, it lets you avoid dealing with minor issues that have no effect on service.
Root cause analysis
Splitting IT monitoring among various tools makes it hard to identify root causes and the IT services that may be affected. What’s more, it adds to the number of people involved in event resolution, with too many information silos covering too many narrow infrastructure segments. This approach is almost worthless at the enterprise level, because there’s no single view of how the infrastructure events are affecting system performance.
Instead of this scattered approach, look for root cause analysis of performance and availability issues, You’ll be able to more quickly generate a prioritized ranking of service events to help isolate events and figure out the impact each event may have on service levels.
Open support for existing devices
We’ve seen now that silos of IT monitoring only makes service provisioning harder. Legacy monitoring tools only make the problem worse, and more expensive. If you were able to remove even one legacy framework, the maintenance licensing cost savings could allow you to purchase a modern solution and add operations staff to ensure more reliable levels of service. The important thing to keep in mind here is to look for open and extensible monitoring architecture to keep up with cutting edge new technologies.
A success story
Is this the right way to monitor your IT infrastructure? Consider New Zealand’s 2degrees. Three years after launching, the company already had 21 percent of the mobile device market.
2degrees prides itself on innovative customer-centric services (they were the first in New Zealand to offer mobile payment and carryover minutes). Those services, however, add to the company’s network monitoring and maintenance burden.
According to 2degrees’ Ahlers, unified monitoring makes it easy to maintain reliable incident detection and incident analysis. The company uses root cause analysis in reporting, and the company’s network configurations are now speedy and efficient.
Network health and monitoring data enable 2degrees network administrators to visualize risk areas. Its network operations center has a consolidated view of all of its systems and networks. And 2degrees is able to ensure an optimal customer experience as new networking technologies and cloud services come online.
Splitting your networking monitoring into silos is costly and does nothing to give you a look across all lines of business. With unified monitoring, you can be sure your network performs reliably, and the checklist we’ve provided here will make sure you find the solution that works best for your business.
Brian Wilson is Senior VP of Customer Success at Zenoss. He can be reached at [email protected]
Filed Under: Infrastructure