When the network goes down at night

There's a specific quality to a network outage that makes it feel worse than other IT failures: it takes everything down simultaneously. Email. Cloud applications. POS terminals. VoIP phones. Security cameras. The moment the switch goes dark, the entire operation stops.

For Seattle and Eastside businesses running evening or overnight operations, a network outage at 9 PM isn't an inconvenience — it's an operational emergency with a clock ticking.

This guide is the structured triage process that gets your network back fastest.

The three-question diagnosis

A network outage is almost always one of three things: an ISP problem, a hardware failure, or a configuration error. The first three minutes of diagnosis split the problem into one of these paths, and each path has a completely different resolution.

Question 1: Does anything have internet connectivity?

Test from multiple devices — a laptop, a phone on Wi-Fi (turn off cellular data first), a desktop. Try different spots in the building. If any device has internet access, the ISP is not the problem.

Question 2: Can devices communicate with each other even if they can't reach the internet?

Try to ping or access a network resource (a printer, a NAS, another machine) from an affected device. If local network communication works but internet doesn't, the problem is at the router/gateway — not the switches or cabling.

Question 3: Is the problem in one area of the building or everywhere?

A zone-specific outage almost always points to a switch or the cable connecting that zone. A building-wide outage with working core network points to the router/firewall. A complete outage with no local connectivity points to the core switch.

ISP vs. hardware — how to tell the difference fast

If you suspect the ISP:

1. Check the modem/ONT lights. The modem or ONT (if you have fiber) has status LEDs. A blinking or off "online" light, or a solid red indicator, means the ISP circuit is down. Look for: - Cisco/Arris/DOCSIS modems: "Online" or "DS/US" LEDs - Frontier/CenturyLink/Ziply fiber ONT: look for a "LOS" (Loss of Signal) indicator - Comcast Business modem: power and status lights

2. Bypass your router. Connect a laptop directly to the modem with an ethernet cable. If the laptop gets internet access directly from the modem but not through the router, the router is the issue, not the ISP. If the laptop can't get connectivity directly from the modem either, it's the ISP.

3. Check the ISP's outage map. Comcast Business, CenturyLink/Quantum Fiber, Frontier, and most carriers have outage maps on their websites. A known area outage in your zip code is the answer.

4. Call ISP business support, not residential. Business accounts have different support numbers with better SLAs. Find your carrier's business support number — it's on your bill — not the number on their homepage (which routes to residential). Be ready with your account number.

If the ISP circuit shows as connected but you still have no internet:

The problem is somewhere in your local network stack. Work from the outside in:

Router/firewall first: Can you access the router's admin interface from a wired connection? (Usually 192.168.1.1, 192.168.0.1, 10.0.0.1, or whatever your default gateway is.) If you can reach the admin interface, check: - WAN connection status — does it show a public IP address assigned? - DHCP pool — is it full? (If all IP addresses in your range are assigned, new devices can't get connected) - Firewall logs for blocked traffic

Core switch second: Are all the LED indicators on your managed switch showing port activity? A switch that lost power, overheated, or had a firmware crash will have all ports dark. Managed switches with a console port can show error messages when accessed serially.

Physical layer last: Ethernet cables fail more often than people expect, especially in environments with heavy foot traffic, chair movement, or temperature cycling. A damaged cable at the patch panel connection or at the wall jack can take down one or many ports depending on where the damage is.

The equipment inventory problem

Here's what slows down almost every after-hours network outage for small businesses: nobody can find the documentation.

Where is the router? Where is the main switch? What's the admin password? What ISP are we using and what's the account number? Is that box in the server closet the modem or the router? Why are there two routers?

This information typically lives in the head of the person who set up the network — who is not available at 10 PM.

If you're reading this before an outage, this is your action item: create a single document (paper or digital) that contains:

  • Network equipment list with make/model and location
  • Default gateway IP (the router's internal address)
  • Admin URL/credentials for router and managed switches
  • ISP carrier name and business support phone number
  • ISP account number
  • Router WAN IP (static, if you have a static IP)
  • Any VLAN configuration
  • Date of last configuration backup

Put this document somewhere accessible to the person who will be dealing with the 10 PM failure — not just on the server that just went down.

Seattle-area ISP escalation: what actually works

Comcast Business

  • Business support: 1-800-391-3000 (24/7)
  • Outage map: businessoutageinfo.comcast.com
  • SLA claims: Business accounts with Ethernet service have uptime SLAs — if the outage exceeds your SLA threshold, request a credit proactively

Quantum Fiber / CenturyLink Business

  • Business support: 1-877-453-8353 (24/7)
  • Outage map: available at quantumfiber.com
  • Note: Quantum Fiber has expanded significantly in the Seattle area; if you have recently switched, verify which support number corresponds to your new service

Ziply Fiber (Northwest coverage)

  • Business support: 1-866-699-4759
  • Ziply has good coverage in the Eastside markets (Bellevue, Kirkland, Redmond, Issaquah)
  • Business customers have priority queue access vs. residential

AT&T Business (Enterprise/dedicated circuits)

  • Business support: 1-800-321-2000
  • If you have a dedicated circuit (not broadband), escalate immediately to NOC level — dedicated circuits have 4-hour hardware dispatch SLAs in most business agreements

When calling the ISP after hours:

  • Have your account number, service address, and contact phone number ready
  • Ask specifically: "Is there a known outage affecting my address?" — they can look this up in 30 seconds
  • If no outage is listed: ask for a Level 2 tech or "Network Operations Center" — front-line after-hours support is often limited in what they can do
  • Request a ticket number and name of the agent — this matters if you need to escalate further

Hardware failures: the three most common after-hours scenarios

1. Router or firewall crash

Routers and firewalls fail in two ways: hard failure (won't power back up) and soft failure (running but not routing). A soft-failed router often has lights that look normal but passes no traffic.

First response: A power cycle — unplug for 30 seconds, plug back in — resolves a substantial percentage of soft router failures. If the router has a reset button, do NOT press and hold it (that's a factory reset) — just the power cycle.

If power cycling doesn't work: Access the console. Most business routers (Cisco, Meraki, Ubiquiti, SonicWall, Fortinet) have a USB or serial console port that allows direct access even when the management interface isn't responding.

If the router is completely dead: You need a replacement or a failover. If you have a cell-enabled backup router or a consumer router as a backup, this is when it gets connected.

2. Core switch failure

A switch failure is distinguishable from a router failure: with a dead core switch, you lose both internet AND local network connectivity. Devices can't talk to each other.

Power cycling a failed switch resolves most soft failures. Hard switch failures require replacement hardware.

For managed switches: before assuming hardware failure, check whether the switch has entered a protection mode or loop-prevention mode (Spanning Tree changes, broadcast storm) that is blocking all ports. This is a software-level failure that looks like hardware.

3. PoE overload

Many small business networks use Power over Ethernet (PoE) switches to power access points, VoIP phones, and cameras. PoE switches have a total power budget. When the budget is exceeded — often after adding a new camera or replacing an access point with a higher-wattage model — the switch begins shutting down PoE ports in priority order.

Signs: Specific devices (access points, cameras, phones) going offline rather than all devices. The pattern often follows the PoE priority order set in the switch configuration.

Fix: Reduce PoE load (disconnect lower-priority devices temporarily), or replace with a higher-wattage PoE switch.

When remote troubleshooting isn't enough

Most network outages can be diagnosed and resolved remotely — with one catch: if the problem is physical (bad cable, dead switch, power issue), someone needs to be on-site.

Remote engineers can: - Diagnose whether the problem is ISP or hardware - Access managed switches and routers that have out-of-band management - Walk a non-technical person through physical triage steps - Identify the specific piece of hardware that needs replacement

But they cannot: - Swap a cable that physically needs swapping - Replace failed hardware - Press the power button on equipment that needs a power cycle

For businesses in Seattle and the Eastside, this is the distinction between a remote-only after-hours provider and one with genuine local presence. A provider who can be on-site in 45–60 minutes when physical work is needed changes the outcome of roughly one in five network outages.

Network resilience: what a $200–$500 investment buys you

The most cost-effective resilience investment for small businesses isn't redundant hardware everywhere — it's a failover path for the most critical function: internet connectivity.

4G/5G cellular failover router: A cellular backup router ($150–$400 hardware, $30–80/month for a business cellular plan) provides automatic failover when your primary ISP circuit fails. The router monitors the primary connection and switches to cellular in 30–90 seconds, transparently. For most small businesses, this single investment eliminates 80% of the operational impact of ISP outages.

Popular options for small business: Pepwave MAX series, Cradlepoint E100, Digi EX15. These are configured in advance for automatic failover and require no manual intervention during an outage.

What this doesn't solve: Hardware failures on your internal network (failed core switch, router crash). Those still require intervention. But ISP-caused outages — which represent the majority of small business network outages — become self-resolving.

Before the next outage: a 30-minute preparation checklist

  • [ ] Know your ISP's business support phone number and have it saved in your phone (not in your email, which you may not be able to access during an outage)
  • [ ] Document the physical location of every network device: modem, router, and switches
  • [ ] Confirm default gateway IP and admin credentials are documented and accessible offline
  • [ ] Check router firmware — most small business routers have a "check for updates" option in the admin interface; out-of-date firmware is a common cause of instability
  • [ ] Consider a cellular failover router if network uptime is critical to your operations
  • [ ] Save an after-hours IT support number before you need it — finding technical help during a live outage adds 20–30 minutes to recovery time