Power Outages can happen at any time, and if your business relies on IT uptime you will need to plan for when they do. Proper planning for a power Outage will benefit your, your team, your users and your business greatly, as trying to react to a power outage is likely to do the opposite.

At Enviromon we have made it our mission to help IT professionals and businesses owners like you save money and prevent avoidable disasters from environmental factors that threaten your critical business assets.

This is why we have put together this Server room power outage checklist to give you some advice on planning and managing for power outages in your server room.

Server Room Power Outage Checklist

Each company's server mix and network infrastructure will be different, but this list should give you an idea of how to structure a server room power outage procedure to organize a plan of your own.

1) Make sure your Server Room is monitored.

If the power goes out at your facility, a quick response time will be a factor in minimizing the damage.

√ - Voltage sensor alerting - Being notified that a power failure has occurred is critical. An undetected power outage can have devastating affects on your network devices and users. 

Voltage monitoring sensors will detect when the voltage has ceased and will send email or SMS alerts to key personnel to notify them to take action. This can be critical to detecting the outage as soon as it happens, including after hours or at remote sites when there isn’t always staff around to notice.

√ - Temperature sensors alerts - If the power has gone out from your facility, this means that the A/C unit for your server room is off as well.

This becomes a big problem if your servers are still running during the outage off of a UPS backup battery supply. Heat will still be generated by the devices without sufficient cooling. Temperature monitoring sensors that are powered by back-up power as well alerted you to the rising temperatures in your server room. This can give the needed warning, that you will need to begin to shut down before the UPS power is exhausted to prevent hardware damage.

2) Make a plan for your UPS battery back-up system

If you don’t have a UPS, or uninterruptable power supply you should invest in one. If you do have UPSs for your server room it is best to make a plan of action for eventual power outages. Remember in the event of a power outage you may not have a computer to use to recall your plan, make sure to have a printed copy around just in case.

√ - Make a list of what your UPSs cover and what they don’t.

Typically only your critical servers and network devices should be connected to the UPS’s back-up power supply, along with your environmental monitoring system. Having a headless server on UPS is great, but if the switch isn't on the UPS too you won’t be able to remote into it to shut it down.

√ - Know what needs to be shut down immediately and know what isn't critical to shut down before the UPS dies because it will be fine regardless.

√ - Test your UPC battery at least once a month if power outages are common. Or at least every 3 months to ensure that they are in working condition when needed.

3) Make a full list of the proper boot order for all servers and devices in your server room.

Again this is something to print out, as you may not have access to a digital copy of your plan.

√ - Typically you want to start up your devices from the "outside in," meaning modems first, then routers, then switches, then servers and appliances.

√ - If there are large amount in-house applications, this should also be documented (example: a domain controller might need to be completely booted before the exchange server comes back on, etc.)

4) Be prepared with the proper safety gear.

√ - Make sure your server room has flashlights and some sort of emergency lighting solution, as this will allow you and your team to be safe while working in the dark.

5) Make a “test” checklist to make sure everything is a back-up and running.

Bringing your network back online will be only one-half of the battle, the other half is to test that everything came back on and is reconnected to the network.

√ - Make a checklist of all of your systems and how they will need to be tested to make sure everything is back up and connected properly.

Your list might include: sending test emails, pinging a remote site over the VPN, checking that your workstation can talk to the domain controller, check fileserver access, print a test page to every printer in the building, walk the floor and make sure all the workstations, monitors, network printers, and etc. have come back on.

6) When everything is confirmed to be working, inform your users.

√ - Send a brief email to your users to let them know everything is back up and running and available.  

This will not only reassure them that everything is under control, but it also gives them a clear point of reference so if they are having technical issues, they know it isn't because the systems are still down and they need to call it in.