Service Availability Overview

In a conventional network, router outages can occur because of denial of service (DoS) attacks, line module failure, switch route processor module failure, software defects, feature upgrades, or complete router failure. These outages result in subscriber downtime.

To reduce subscriber downtime, a network must have the following capabilities:

JunosE Software uses a multi-layered service availability approach that enables you to provide uninterrupted delivery of services with the help of reliable, highly available, and redundant hardware and software components.

Figure 1 illustrates the multiple layers of JunosE Software service availability.

Figure 1: JunosE Software Service Availability Layers

JunosE Software Service Availability Layers

The security layer protects the network from DoS attacks.

The network resiliency layer protects against port, link, and node failures. You can configure IEEE 802.3 ad link aggregation for Ethernet, and Virtual Router Redundancy Protocol (VRRP) to improve network resiliency.

The software availability layer protects against software failures by using hot-fixes or installing a higher-numbered software release. You can perform a unified in-service software upgrade (ISSU) instead of the conventional software upgrade to reduce outage. You can eliminate or reduce single points of failure by configuring stateful SRP switchover (high availability). Any network component with an uptime of 99.999 percent is considered highly available with a downtime of less than 5 minutes in a year.

The hardware redundancy and design layer introduces redundancy in the network in the form of multiple power supplies, cooling devices, line modules, and sometimes even a router. For instance, you can install a backup line module in your router to protect against line module failure. You can also configure a router as a backup router that accepts subscriber login requests when the master router fails.

Service Availability Versus High Availability

High availability is a measure of the uptime of a network or network component. A network component that has a downtime of 5 minutes is accessible or available 99 percent of the time. If a failure occurs, a backup component is available within 5 minutes. A highly available network is a network that has components that either have high reliability or have the ability to recover very quickly from a failure, or both.

Service availability refers to the ability to provide uninterrupted delivery of services. For example, from the time when a component fails to the time when the backup component is accessible, the delivery of services is interrupted. To provide uninterrupted delivery of services, highly available components must maintain session details and other data across failures. Service availability can thus be defined as the ability to provide uninterrupted delivery of services using a highly available network.

Related Documentation