7 Common Server Room Problems for Businesses to Consider
Sometimes it takes a server room disaster to prompt data management changes in an organization. It is undoubtedly preferable that businesses take appropriate measures to protect their server rooms even before disaster strikes, but excessive optimism, and the inertia of day-to-day operations, makes it easy to forget that the dangers are real.
This is why so many companies only take measures to adequately ensure stack safety after their local server rooms are impaired through natural or predictable causes – or after news of a similar event at a different company reaches the ears of those who make decisions about server maintenance.
Common Problems With On-Site Server Rooms
Before attempting to determine how best to establish the most secure yet cost-effective means of storing your data, it’s important to recognize that local server rooms are not inherently disasters waiting to happen. Rather, there are specific common sources of damage that can plague your data that you should make sure to protect. You decide you’re better off migrating your stacks to a shared offsite facility, or you may prefer to keep them in-house – either way, when you address the specific server facility threats that most often destroy data assets, you’re able to dramatically minimize risk.
Let’s take a look at some basic best practices for precautions that will help to keep your on-premises data safe and accessible for the needs of your business.
1. Inadequate Temperature Control
To maintain a secure space for your local servers, you must create the proper physical conditions for them.
That includes an area that has proper temperature control. The air surrounding your hardware needs to be in a range that keeps heat from damaging your servers. Typically this is a range of 68 to 72 degrees. Not only must you account for temperature in the server room, but also impact on temperature from outside the room... such as direct sunlight or other factors that may raise temperature. Your servers will generate heat too, and the environment will need cooling that can account for that. Ideally, you need to make sure you’re able to control the room temperature separately from the rest of your facility. Countless servers have been lost to too warm of an environment.
2. Insufficient Ventilation
Similarly, you will want to make sure that there is sufficient space surrounding your equipment for air to circulate. How much of a gap you need depends on the specific parameters of your machinery, but all heat-generating data storage devices (and they all generate heat) require air circulation to make sure they don’t overheat – poor ventilation necessarily results in inadequate cooling.
Therefore, if the space you use for your servers is small enough that you need to stack them, make sure to use dedicated server racks, rather than just piling servers up on top of each other.
3. Imbalanced Moisture Levels
Keep in mind that temperature is not the only environmental condition concern that impacts the location and well-being of your servers. Controlled humidity can be at least as important. High humidity can result in rust, corrosion, short-circuiting and even the growth of fungus that can attack the machinery. Too little moisture in the air is also a concern, as an exceedingly dry environment can result in electrostatic discharge, which in turn can cause system malfunction and damage.
Protect your data from more particular sources of moisture as well – determine where the pipes in your building are and make sure that they cannot spring a leak on your servers.
4. Too Much Jostling
Other potential environmental hazards include vibrations that can disturb a rapidly spinning hard drive, or dislodge the boards and chips that make your technology run. Even tiny scratches, particularly to a hard drive, can obliterate the data you are working so hard to preserve without damage.
Outside of an earthquake, which are not all that common outside of California, vibration is usually the result of moving the server or bumping into it. If servers are too close to hallways or outside walls, they are also more subject to jostling caused by machinery and people passing by – particularly the potential movement caused by a large truck, for example. Plan your cushioning and other stabilizing forces well, and your servers should not be disturbed by vibration.
5. Clutter and Disarray
The server room is yours - keep it as orderly as necessary. Make sure that your cords are untangled and that there’s no risk of tripping over wires. A power distribution unit is likely to help you in this regard.
Note that no matter what you do, you may be subject to a freak accident, like this exploding fire extinguisher.
6. Power Volatility
Among the potential events that threaten your servers are power outages, whether blackouts, spikes or brownouts. Make sure that your system is covered by providing redundant power backups – a standalone generator (or two), for example, or an uninterruptible power supply (UPS) device. Often, in the event of electricity flow volatility, a reboot of the system will return your servers to their functionality, but there is no guarantee against severe damage, so do what you can to prevent power outages and spikes from affecting your servers. The power backup systems should be tested regularly and also monitored. It is best to identify backup system shortcomings during a test rather than during an actual power outage, itself. Amongst the things to look for are depleted batteries in UPS, failure to start with a generator and low fuel or fuel that has been rendered unusable by sediment in the fuel tank.
Power outages are nearly always “acts of God” – the results of extreme weather. Sometimes, however, they are the result of human error at the electric company. Every so often, they happen because of sabotage. No matter what causes your system to fall electrically, if you have a backup system in place, you will preserve your data as needed.
7. Acts of Intentional Malice
Concerns about intentional sabotage should not be overlooked, no matter how unlikely they may seem to your particular situation. Security breaches are always a possibility, and making sure that your servers are protected against both physical and cyber-intruders is critical.
If a hostile party steals your data or manages to install viral code into your system, you will not be able to conduct business as usual – to say the least. Affix security cameras to capture anyone entering your server room without authorization. Ideally, the data recorded by camera should be archived for retrieval and viewing in the event of a breach. Install locks – the kind that require a key or a combination, in the server room itself and also on any server racks that you may have installed. Secure your phone room and fiber/copper conduit that comes into your building/property.
Consider Your Options Soberly and Methodically
When you maintain your own local server room, you have all of the privileges and benefits of internal control, but you’re also charged with protecting the room from any of the forces that might cause damage to your servers. That said, colocation services and a shared datacenter may be better for you, exempting you from much of the burden involved with alleviating these risks. Speak to an expert to find out whether your needs might actually be best served by an alternative to your own on-premises server room.
Factors to consider include all of the risks to servers that are mentioned above, as well as the cost of equipment, and the benefit of turning your delicate system over to those who may be in the best position to protect it from damage. Only one who understands the pros and cons of keeping your servers in-house as compared to farming the system out, as it were, can help you determine which approach is best for you and your business.