Summer is here: Help your servers beat the heat!

Summer is here: Help your servers beat the heat!

Written by Jame Lijewski, Network technician & Support Center

Summer is here, and so is the heat!

Servers and other network equipment gets hot, but how hot is too hot? Those high temperatures and other environmental conditions can threaten your business productivity as much as security threats. Some main environmental conditions that are a cause for concern are high temperatures, humidity, water damage, and physical security.

There are a few key things you can do to prevent environmental damage to your vital IT infrastructure that can minimize downtime for your company.

Server & equipment monitoring

Servers & other vital equipment usually have environmental monitoring. Some alerting may be able to be configured for CPU temperature, fan speed, etc.

Server room temperature monitoring

The temperature of your server room should be consistently between 64-80°F. Due to the vital nature of the equipment, it is usually recommended to get a separate AC unit just for your server room. In addition to monitoring the temperature of the room, it is also best to monitor temperature near the AC unit to detect if it’s still operating efficiently, as well as the air coming into the server rack.

Server room humidity monitoring

The humidity of your server room is just as crucial to monitoring as the high temperatures. When the room is too dry, it can cause buildup of static electricity of devices. On the other hand, when it’s too humid, corrosion can wreak havoc and condensation can build up on electronic devices, causing premature equipment failure.

Flood monitoring

Most leaks or flooding in server rooms are caused by leaking air conditioning units. Water sensors are best placed under or near the AC unit for the server room, and when possible use a rope sensor around the room or server racks.

Physical security monitoring

In terms of physical security there are many options to meet your basic needs. You can go for a more basic route from something as simple as a door opening sensor, that can be used to alert unauthorized access to the room after hours. Another more secure option would be installing security cameras to record key infrastructure, pointing the cameras directly at the server room access door and/or vital equipment.

As with all monitoring, it doesn’t do much good unless you’re setting up some type of detection and response. Depending on your organization, there may be different individuals you will want to notify such as the main IT contact, CEO, or other key individuals. Not only should the individuals be notified, but you will want to have a clearly documented plan of action. Here are some questions you may want to ask yourself about your own disaster recovery plan:

  1. If the primary contact is on vacation or on a temporary leave, does your monitoring notify other individual(s) as a fail safe?
  2. Do you have a plan of action for each type of incident? For example, if the AC unit fails, have you previously discussed an emergency HVAC technician contact with your IT department and given them authorization to get HVAC involved?
  3. Does the monitoring and alerting work? When was the last time the monitoring was tested to ensure the alerting is working properly? This is especially important to verify after any recent email hosting changes.
  4. Do you have any environmental warning configured from your electric company? Many now have the ability for email alerting if there are known power issues in your service area. On that note, do you have a backup power supply to gracefully shut down your servers, or have a backup generator to keep your business running until service is restored?

Disaster recovery plans are important. Be sure to have a copy on-site as well as off-site due to accessibility reasons. If you are trying to review your disaster recovery plan on a server that has some environmental problems, your hands could be tied unless you have a physical copy elsewhere or a copy of it in the cloud.