Keeping Your Cool When You Can’t Keep Your Cooling

The last thing you want to happen in the middle of a sweaty New York summer is to have your air conditioner fail. When this happened to me recently, the worst case scenario was a few uncomfortable weeks and a very unhappy wife. 

But while the lack of air con may be uncomfortable and inconvenient for me, it can be a real problem for firms that rely on high-performance technologies, utilizing racks of expensive servers packed tightly into datacenters where location and space is everything.

If trading firms lose cooling in their on-site server room or at the third-party datacenter that hosts their infrastructure, systems can rapidly overheat, forcing them to stop trading all together. This can result in missed trading opportunities, or worse, it could slow their systems, resulting in bad trades. 

Most firms that depend on performance know well enough what to do to maintain it, and have plenty of safeguards in place—for example, backup cooling systems to prevent hardware running too hot in case their air con fails, and operational networks that monitor everything from temperature to server utilization for the slightest hint that anything may be about to go wrong. Some even utilize self-diagnosing and self-healing infrastructures to ensure they not only keep running, but also maintain optimal performance regardless of what happens. 

One factor is third-party datacenters, where in the event of a server failure while a trading firm’s data or IT guy is on vacation, the replacement will be handled seamlessly by the datacenter operator as part of its service. 

The threat of an imminent outage isn’t just a technical issue: It’s also an administrative one. A contract coming due for renewal shouldn’t take anyone by surprise.

Another is that in the event of an issue affecting the datacenter as a whole, a firm may lose out on opportunities, but can feel relatively secure that its peers are all in the same boat and are likely unable to profit from its misfortune.

The threat of an imminent outage isn’t just a technical issue, it’s also an administrative one. For example, a contract coming due for renewal while a market data manager is away shouldn’t take anyone by surprise: If the contract isn’t being renewed, end users should be aware of any changes and have alternatives in place. 

If the contract is being renewed, any negotiations should not only have already been resolved, but everything should be set to seamlessly roll over without any interruption to service, any additions should be tested and ready to come online, and those who sign off on market data costs for their business line should be aware of—and happy with—any changes to what they’ll pay so that the next round of invoices are handles appropriately. And any inventory management platform should be set up to allocate and process those costs, whether they relate to regular market data spend, related non-data costs, or soft commission payments—for more detail on each of those, check out recent stories about new capabilities being rolled out by The Roberts Group and Screen Infomatch in sibling publication Inside Market Data.

Don’t Come Back

Of course, the last thing you want to happen while you’re away is to be told not to come back: that you’ve ensured everything runs so smoothly and effectively in your absence that your services are no longer needed. 

And one of the keys to reminding people how essential you are is to see yourself not as a cog in the machine that can be easily replaced, but someone providing a service that revolves around knowledge gained over your whole career, and the expertise to put in place processes that ensure everything will work smoothly without you. But just because you’re eliminating “key man risk” doesn’t mean your firm should consider eliminating their key men (and women). 

Because if they’re doing their job correctly, a firm’s data professionals should be as strategic an asset as the data itself.