On The Move
At YABBIT we are all ears. You spoke, and we listened, and created an all-in-one, cloud-based communication system for your business.
HIGH AVAILABILITY TO THE NTH DEGREE
Commercial efforts to use IP to transmit voice calls began in the early ’90s. However, adoption of VoIP within telephony markets was limited as result of a variety of technical difficulties. However, each improvement increased acceptability, and today, VoIP has become as ubiquitous as switched-circuit telephones.
IP-based voice networks are growing rapidly thanks to a number of factors including their cost-effectiveness and the opportunity for businesses to increase productivity through easy information sharing, maximized utility of available bandwidth and optimized converged network services.
As service providers continue to offer rich IP telephony services to subscribers, the need for high availability (HA) is becoming critical. HA is not a precise technology, but a goal based on specific business needs. HA improves reliability by reducing downtime, minimizing outages, and limiting the impact of outages if they do occur. Accordingly, redundancy is indispensable for achieving HA and providing uninterrupted 24/7/365 service.
In this document, we examine the importance of HA and redundancy in today’s IP-enabled voice networks while describing how the YABBIT Platforms n-Share empowers service providers to withstand failovers, maximize uptime, and offer new competitive positioning to their IP telephony subscribers.
HIGH AVAILABILITY
refers to the ability to communicate with anyone, anywhere, at any time. For enterprises relying on IP-based communications, the absence of HA can result in decreased productivity, lost revenue, customer dissatisfaction, and weakened market position.
Like other real-time applications, IP-based voice applications are extremely bandwidth and delay-sensitive. For IP-enabled voice networks to be a realistic replacement for standard public-switched telephone network (PSTN) telephony services, customers must receive the same high level of voice quality they can obtain with basic telephone services. For voice transmissions to be intelligible to the receiver, this means voice packets must be guaranteed certain compensating bandwidth, latency, and jitter requirements to avoid dropped calls or excessive audio delays (otherwise known as jitter). HA ensures that VoIP voice packets receive the preferential treatment they require. To achieve HA, users must be able to place and receive calls during peak-load call rates, planned device maintenance, or unexpected failures.
CAPACITY
a measurement of the volume of traffic a network is engineered to handle. Typically, voicenetworks are engineered to handle a target peak-load capacity measured in calls per second (CPS). Target peak-load capacities are specific to each business and industry, and are based on measured busy-hour call rates. Traffic between November and Christmas, for example, might be the target peak-load capacity for a mail-order company’s IP voice network.
REDUNDANCY
Reliability, resiliency, and availability are sometimes used interchangeably when discussing HA. While all three terms are related to the concept of HA, there are significant differences:
| AVAILABILITY | DOWNTIME PER YEAR | TYPES OF SYSTEMS |
|---|---|---|
| 90.0000% | 36 days, 12 hours | N/A |
| 99.0000% | 87 hours, 36 minutes | General PCs,Fax Machines, Printers |
| 99.9000% | 8 hours, 46 minutes | ISPs & Non Critical Business Systems |
| 99.9900% | 52 minutes, 15 seconds | Data Centers |
| 99.9990% | 5 minutes, 15 seconds | Redundant Storage Systems |
| 99.9999% | 31.5 seconds | Military Defence Systems |
THE FOLLOWING TABLE SHOWS WAYS IN WHICH AVAILABILITYCAN BE EXPRESSED:
SOURCE
Infonetics Research, Comdisco/BellSouth/Oracle Vulnerability,ContingencyPlanning Research In February 2012, the United States FCC made new rules that
require VoIP providers to report 30-minute outages that affect 911. Simply doing routine upgrades could create this kind of outage in many networks. Source: “MakingVoIP Geo-Redundancy Actually Work Well” by Mark RLindsey REVENUE PROTECTION Service providers often overlook investing in solutions designed to address reliability, resiliency, and availability because it has traditionally been difficult to highlight the direct link between HA and revenue generation.
However, by quantifying the real cost of IT-induced downtime and evaluating the business-continuity options available, providers can avoid the risk of downtime by matching their spending on protection solutions with what the business stands to lose.
In 2016, network downtime cost US businesses an estimated total of $700 billion. Today, enterprises of all sizes face ever-growing costs of unplanned downtime from natural disasters, system failures and cyberattacks.
In its 2016 report, the Ponemon Institute found that the average cost of IT downtime is $9,000 per minute. A recent Gartner report added context observing that while small enterprises may face lower downtime costs of around $140,000 per hour, medium and larger size operations can face anywhere from $300,000 to over $540,000 per hour.
Even a short disruption can have a significant impact on retail outlets and other businesses that operate in an increasingly cashless economy
| YEARLY COST MATRIX | BEST-IN-CLASS | INDUSTRY AVERAGE | LAGGARDS |
|---|---|---|---|
| Business-Interruption Events | 3 | 2.3 | 4.4 |
| Time Per Business Interruption | 1 | 1 | 9 |
| Total Disruption (hours) | 03 | 2.3 | 39.6 |
| Average Cost Per Disruption | $101.600 | $181.770 | $99.150 |
| Total Cost of Business | $3.048 | $418.071 | $3,926.340 |
TECHNOLOGY
YABBIT n-Share technology addresses the requirements of reliability, resiliency, and availability by providing a framework for empowering service continuity, system expandability, and ease of maintenance.
The following table compares the cost of downtime for businesses with varying levels of preparedness: n-Share uses redundancy to ensure that call-processing systems are sufficiently reliable and scalable to handle the required number of users and devices and resilient enough to handle various network and application outages or failures. n-Share runs on separate general-purpose Linux machines (or nodes) created and managed specifically for, or dedicated to, the n-Share application. n-Share is deployed as an active-active redundant application. This means that all nodes are online and pass call information, feature codes, voice mail, and other traffic simultaneously under normal conditions. If one node fails or loses connectivity, traffic fails over to the remaining n-Share nodes seamlessly, with no down time or loss of network connections. n-Share is engineered to scale in a virtually limitless fashion while still being managed as a single application. It can be deployed as a 2-node configuration or in multi-node clusters. n-Share nodes support geographically-dispersed multi-box redundancy ,making them ideal for disaster-recovery scenarios. A data center with locations on the east and west coasts, for example, might install an n-Share node at each location to spread active communications across datacenters and reduce impact should disaster strike one of the data centers. Global enterprises, on the other hand, might deploy n-Share nodes around the world, according to region or lines of business (for example, one for North America, a second for APAC, and a third for EMEA). This model enables an enterprise to address HA and redundancy across differing territorial boundaries
EFFECTIVE RESOURCE UTILIZATION
active clustering actively forwards traffic to achieve maximum performance gain.
LOAD SHARING
in an n-Share deployment, the call processing load can be shared among multiple nodes for faster response times.
HIGH REDUNDANCY
Active operation enables each n-Share node to provide redundancy to all other nodes. In a hypothetical situation where all nodes in a cluster go down except one, the traffic is still processed (up to the capacity of the single node), thereby indicating the high level of redundancy available through n-Share.
SCALABILITY
fail over time is the quintessential index by which to judge the quality of an HA system. The faster the fail over, the less impact there will be to business continuity. With n-Share, fail over is as seamless as possible in order to maintain service and deliver the required user experience
FAST STATEFUL FAILOVER
fail over time is the quintessential index by which to judge the quality of an HA system. The faster the fail over, the less impact there will be to business continuity. With n-Share, fail over is as seamless as possible in order to maintain service and deliver the required user experience.
PERFORM MAINTENANCE WITH ZERO DOWNTIME
upgrading and patching hardware and software means going offline for a period of time. With n-Share, system maintenance can be performed any time it is required, without having to plan for downtime. Minimizing downtime dramatically improves business continuity, without burdening IT, impacting performance, or affecting user productivity.
NO SINGLE POINT OF FAILURE
in an n-Share topology, at least one alternative path for traffic flow is always available, providing the highest levels of continuity and availability
PROTECTION FROM DDOS ATTACKS
n-Share’s ability to replicate services across nodes offers highly effective,fault-tolerant protection against high-volume attacks designed to bring down critical business applications and resources.
DISASTER RECOVERY
n-Share protects against site-level failures, from natural disasters to man-made activities, such as configuration errors and overwhelming service demand. You can even test your disaster-recovery scenarios in a non-disruptive manner.
CONCLUSION
GOT MORE TO YABBER ABOUT?
Get in touch with us.
