Opening Cross: ‘Key’ Issues: Major to Minor, Minor to Major

I’m as fond of musical analogies as I am of talking about low latency in the language of Formula One racing, or using mixology or winemaking to symbolize the complex blend of data sources and services that exists in any modern trading environment. But I could use an orchestra of instruments to achieve the same analogy, or talk about how partnerships between companies—such as Victory Networks providing market data to power a new real-time options analytics calculator built by Options Research (formerly ORATS), or a similar deal whereby Morningstar is providing fund data to Style Research to support the latter’s Peer Insights fund comparison tool—blend melody and harmony into a cohesive piece of music, or I could compare major and minor keys to positive or negative events (depending on whether something major or minor is good or bad in a particular circumstance).

For example, an incident last week that turned out to be minor but could very easily have been major was an outage in a power supply module at Equinix’s LD4 datacenter in Slough, UK, that hosts major infrastructure operators, marketplaces and end-user firms, who rely on it for connectivity not just to the markets hosted within, but also for access via third-party networks to other marketplaces hosted in other datacenters—such as the wireless network connectivity that microwave network operator McKay Brothers is now providing between Euronext’s co-location facility in Basildon, Essex and other London-area datacenters, including LD4 on the opposite side of the city.

An Equinix spokesperson says the issue only affected four clients in the datacenter, though the fact that Equinix’s online management portal was unavailable during the incident and that one affected company’s resiliency procedures didn’t work left the company initially wondering if it was a bigger incident. Indeed, in some cases, firms prefer to hear that an outage is a major, industry-wide incident rather than just something affecting them, because—like the uncharacteristic major Bloomberg outage of just over a year ago, or, for example, when seismic activity severs an undersea cable serving an entire region or market center—when everyone suffers the same problem, you effectively have a level playing field, whereas an outage that only affects part of the market hands a huge advantage to those still connected and functioning as normal.

That the Equinix “outage” affected such a small number of clients is a testament to the processes and tools put in place by those who operate datacenters, networks and mission-critical software platforms to monitor for the slightest issue that might indicate a larger problem, and which ensure minor problems don’t fester or escalate into major ones, while potentially major issues can be isolated and quickly reduced to minor incidents. And that can be more effective than the trained musician’s ear at detecting problems: the musician can hear when an instrument is out of tune, but in this analogy, today’s monitoring equipment would be able to tell that the tension on a guitar string has slackened and is therefore out of tune before anyone even plucks the string.

So next time you’re singing along to Taylor Swift in the shower (we all know you do), spare a thought for these seemingly minor marvels that prevent major issues.