Home › Resources › Avoiding points of failure
  Bookmark and Share

Avoiding strategic points of failure in bandwidth management

points of failure in bandwidth managementWhen preparing or upgrading an organization’s WAN capabilities, points of failure need to be addressed to proactively ensure maximum uptime for the organization.   The Elfiq White Paper on Business Continuity offers more details on this topic.

The following list provides some examples of potential points of failure which can be strategically eliminated:

 

Point of failure Description Resolution
Lack of ISP diversification When single-sourcing to a provider with multiple links, like two DSLs or bonded T1s, the redundancy is limited because the ISP links all use the same infrastructure, sharing the same potential problems and thus all failing at the same time. Combining multiple ISPs from multiple carriers will eliminate this strategic point of failure through diversity.
Lack of technology diversification Combined multiple ISP links using the same technology (T1s per example), they all share the same telecommunications infrastructure in most cases regardless of whether if the service is coming from different ISPs. This setup is vulnerable to the “first 100 yards’’ where the local telco’s infrastructure is the risk factor. Should the Central Office/Point of Presence (CO or POP) no longer be able to operate, all the ISPs connecting to this POP will be unable to transmit data. The first way of resolving this issue is to have the ISPs connect to different COs, even the same technology is used. A second way to solve this problem is to diversify the ISP technologies used at the site, where different COss will be utilized, such as for example, DSL and cable modems.
Utility interventions When utilities are doing work in urban areas, they must have the layout of all the underground wiring, but sometimes cables get damaged regardless. Wireless ISP links are the best and easiest way to ensure connectivity when cables get damaged. Also, using ISPs with different COs will most likely use different underground conduits, reducing the risk of losing bandwidth completely.
Equipment failures When a router or modem fails to operate normally, it will cause the ISP link to halt operation. Also, with the use of link balancers, should a single unit be in use,  it causes a single point of failure. Should the business continuity device fail, all connectivity is lost and manual intervention is required to restore bandwidth. With the use of a Link Balancer, multiple routers and modems are used concurrently, so when one fails, at least one other is able to perform its tasks normally. When the troubled link is restored, the link balancer will use the link as before. In the case of Link Balancer failures, there are two avenues to prevent downtime:
  • Budget sensitive environments will want to consider Elfiq’s unique LAN Failsafe feature, which enables a bypass to pass traffic between the primary link and the firewall in case of a power failure.
  • High availability units are recommended for optimal operations: should a unit fail, a second one monitors the first and takes over traffic management until the first unit returns to normal.