Avoid Surprise Bills by Monitoring the Performance of Your Backup Interfaces
We've all heard the stories of the surprise cell phone bill –thousands of dollars in fees for unintended or unknown usage. I recall a recent story on GlobalPost.com about a woman who also carries her two brothers on her plan. Her brothers (both are deaf-mute) communicate primarily through texting and also use their phones to watch videos. When they traveled to Canada they changed to an international plan and during the two week stay, they had sent more than 2,000 texts and downloaded videos. Later that month she received a bill for over $200,000. Fortunately her mobile carrier waived most of the fees.
This type of surprise can happen in the IT world. Did you ever receive a bill you didn't expect? Some companies have. The culprit: backup interfaces that were active and in use without anyone realizing.
Traditionally, backup interfaces stay idle until certain circumstances occur, then they are activated. Backup interfaces can be physical interfaces such as a DSL, ADSL, ISDN Basic Rate Interface (BRI),T1/E1 or an assigned backup dialer interface to be used within a dialer pool. While the primary line is up, the backup interface is placed in standby mode. Once in standby mode, the backup interface is effectively shut down until it is enabled. Any route associated with the backup interface will not appear in the routing table. This at least is the desired outcome.
However, there are situations where backup interfaces are in an up state and the router is configured to or modified to change routing based on the state of the primary interfaces. Most often changes to configurations and manual changes are inherently error prone, and can cause many problems. For example a route table change may inadvertently include the backup interface(s), therefore routing traffic over high cost interfaces, like an ISDN backup or leased lines, that are billed on data rates as opposed to fixed rates. Performance monitoring of these active backup interfaces can give you pro active warning by setting thresholds on data usage or interface states. If you see data usage on a backup interface, and you do not have any alerts telling you the primary interface went down, you want operations to be alerted so they can quickly investigate to determine if that interface should be in use or not.
Nobody wants their first indication that something is not working properly to be their WAN bill at the end of the month or quarter. It may seem like a simple solution, but it could save you thousands of dollars in the end and the shock of an unexpected $200,000 bill.
How has your WAN backup strategy evolved over the years? Are you leveraging newer services from your providers like Ethernet Services? We'd like to hear your thoughts.
Pete Cruz is the Senior Director of Product Mangement and Marketing for SevOne.