Two’s company: Why you need redundancy for critical applications
The word “redundancy” spells bad news for workers, but for systems it spells comfort and reassurance. Even something as unbreakable as a wireless link benefits from redundancy planning – and the more life or mission critical your communications become, the more you must have redundancy.
In the quest for 100% systems’ reliability, one of the toughest paradoxes is that the more reliable you make things, the worse the disaster when they fail. In the early days of computing, when the mainframe or access to it ‘went down’ on a daily basis, everyone just got on with some paperwork until it was restored. If it happens only once or twice a year, however, everyone is caught unawares. But when disaster is unprecedented, you have chaos.
It happened in Queensland, when Cyclone Oswald dropped twenty-four inches of rain over the 2013 Australia Day long weekend. The widespread damage and disruption cost an estimated AUD2.4 billion. Although telecom networks in the region failed, Vertel’s microwave radio links remained largely unaffected, meeting carrier grade performance standards even during the height of the storm. Another example was the 2012 Telstra Exchange outage in Warnambool, Victoria.
Here’s a second tough paradox: the better your corporate network, the more its integrity is threatened by convergence. The Ethernet LAN began in the 1970s as Xerox PARC’s solution to connecting a few computers to a shared printer. By the 1990s it had become every organisation’s data network, but there were other independent networks still running round the site: the telephone network, smoke/fire alarms, access and environmental control systems and, in manufacturing plants, process monitoring and control networks. Bit by bit these are all converging onto the trusted corporate network – reducing the cabling while adding a huge burden of diverse protocols and CoS requirements.
So high availability has become essential to the entire business operation. And if it’s a multi-site operation that needs to ensure availability across a WAN, then the integrity of the data links becomes utterly critical.
That’s why it is so significant in both the Queensland and the Warnambool examples that Vertel’s microwave links remained operational. You can forget those early tales of microwaves being weather dependent: today’s systems are engineered specifically for local weather conditions to achieve very high availability.
That’s not to say that fibre is not the preferred high performance medium in the right conditions – Vertel relies on a protected fibre core, well-designed for optimal redundancy. But when it comes to capillary access links to remote sites, it’s too costly for a single customer to fully protect fibre from damage by flooding, fire, landslide, earthworks, frost, animals or vandalism – none of which are such a threat to microwave.
Adding redundancy to a microwave link is as easy as installing a second antenna pair. Security is ensured by careful placing to reduce common points of failure – a lot simpler than totally re-routing a second fibre to avoid simultaneous damage to both links.
That is why network managers increasingly look to high performance microwave to add redundancy to their existing fibre links: it’s quick to install and not subject to most of the threats that would compromise fibre.
Your network may be more critical than you think. We all understand the vital importance of healthcare and financial systems, but try telling a teacher with a class-full of bumptious adolescents that a network outage is no big deal!
And, when so many previously autonomous services are converging onto your system, you can be sure that someone somewhere will want you to read Vertel’s fact-packed datasheet on Disaster Planning.