Every few winters, the Portland metro gets a reminder that we live where the Gorge wind meets the wet: an ice storm coats everything, branches take out lines, and parts of the metro go dark for anywhere from an hour to most of a week. The good news is that ice storms are the most plannable disaster IT ever faces — they’re forecast days ahead, they follow a known script, and the preparation is mostly a checklist you run in October instead of a scramble you run in January.

Power is the whole game

Nearly everything an ice storm does to your technology, it does through the power cord. So the preparation starts there, in three tiers.

Tier one: ride through the blips. The flickers and one-minute drops that precede a real outage kill unprotected equipment more often than the outage itself. Every server, firewall, switch, and phone system component belongs on a UPS with a battery that’s actually been tested — batteries age out in 3–5 years, and the storm is a bad time to learn yours did.

Tier two: shut down with dignity. When the outage outlasts the battery, servers should power down gracefully and automatically — a configured shutdown, not a hard stop mid-write. This is software, it’s usually included with the UPS, and it’s configured in an afternoon. An astonishing number of businesses own the hardware and skipped the setup.

Tier three: decide about the generator honestly. Most offices don’t need one; the work-from-home fallback below is cheaper and works. The exceptions are real — operations with refrigeration, on-premises production, or server rooms whose temperature matters — and for them, the generator decision deserves the same arithmetic as any other insurance: cost against honestly estimated hours of exposure.

Internet fails separately, and differently

Ice takes down aerial fiber and cable lines on a different schedule than power — sometimes your office has lights and no internet, sometimes the reverse. Two implications. First, a cellular failover connection earns its keep in exactly this weather, with one caveat: it dies with the power unless the modem and firewall are on the UPS too. Second, if your phones are VoIP — most are now — the internet outage is a phone outage, and clients calling a line that rings dead will draw their own conclusions. Call forwarding to a cell should be a documented, tested procedure someone can execute from home, not a support ticket filed mid-storm.

The work-from-home surge is the real plan

Here’s the modern shape of an ice-storm response: the office closes (or the roads do — in this metro, ice on the West Hills and the bridges makes the decision for you), and everyone works from wherever they have power. That only works if it was true before the storm. The self-check: Can everyone reach what they need without being in the building? Do people take laptops home when weather is forecast, or is the standing instruction “grab your machine” part of the storm plan? Does anything critical live only on an office desktop or the server in the closet? If your answer to the last one is yes, that’s less a storm problem than a continuity problem the storm will find.

The quiet failure mode is the person who has power, a laptop, and no idea whether they’re supposed to be working. Which is a communication problem, not a technology one — decide before the season who declares a storm day and on what channel, and make sure that channel doesn’t depend on the office being up.

The October checklist

Run this once, before the first forecast freeze:

  • UPS batteries tested on every server, firewall, switch, and phone component — with the network gear confirmed on battery, not just the servers
  • Graceful shutdown configured and actually tested once
  • Cellular failover verified — pull the primary connection and watch it take over
  • Call-forwarding procedure written down, with someone off-site able to execute it
  • Laptop-goes-home habit established for forecast weather; anything desktop-only identified
  • Backups verified restorable — power events are when marginal hardware dies for good, and the backup you haven’t tested is a hope, not a plan
  • Storm-day communication chain agreed: who decides, who tells everyone, on what channel
  • Provider expectations confirmed: what your IT support’s storm posture is, and how to reach them if your usual channels are down

The bottom line

Ice storms announce themselves, follow a script, and reward an afternoon of October preparation with a January that’s merely inconvenient instead of expensive. Power protection, an internet fallback, a work-from-home posture that’s real, and a communication plan that doesn’t route through the dark building — that’s the whole list. We’ve helped clients through every metro ice event since 1997, and the pattern never changes: the businesses that ran the checklist spend storm week working from their kitchens, and the ones that didn’t spend it in the parking lot, on the phone with us. Our backup and disaster recovery practice exists to put you permanently in the first group.