Earlier this month Atlas struck the Black Hills of South Dakota. 4-8 inches of snow were forecast for the higher elevations (5000+ feet), but here on the foot hills at 3500', we got 31" of snow. It was a wet, heavy snow that snapped power lines and tree limbs. 60+ mph winds made for zero visability and took out a large number of power poles.
Our little datacenter lost utility power Friday evening, and promptly switched to UPS, which had a lifespan of about 2 hours. Power was restored after 85 minutes, but the decision was made to power off all the servers in case we lost power again, with an eye towards starting recovery procedures in a day or two. The data center was restored to full functionality by Sunday noon, even though the businesses didn't re-open until Monday noon.
We have a complete DR plan, so if the outage persisted for another day, we could have resumed operations at a sister site. The key takeaways here were backup validation for off-site replication, lines of communication between Operations and the affected managers, and validated, sequenced shut-down and power-on check-list. I was able to get on-site through the storm thanks to my big 4x4 and coordinate the shutdown and power-on processes. Without being onsite, we would have had some more challenges due to area wide loss of network connectivity.