How to Set Up Internet Failover for Your Office in Pakistan

If you want to know how to set up internet failover before the next cable fault takes your team offline, this week gave you the perfect reason to start. On 18 November 2026 the PTA confirmed a fault on the SEA-ME-WE 5 (SMW5) submarine cable about 11 km off Pakistan’s shore, and users across the country spent days with slow pages, frozen calls and failed uploads.

The PTA said traffic was being rerouted through other international links, but gave no restoration date (The Nation carried the full statement). Cloudflare Radar data cited by The National showed HTTP traffic dropping by as much as 36 per cent. You cannot fix a cable on the seabed. You can make sure your office keeps working when one breaks. This guide walks through a practical failover setup that a small software house, agency or BPO can put in place in a weekend.

What You Need Before You Start

Keep the shopping list simple. Here is what a typical 10 to 40 person office needs:

  • Two internet connections from different providers: ideally one fibre line and one from a provider that uses a different upstream route, plus a 4G or 5G data device as a third, emergency path.
  • A dual-WAN or multi-WAN router: MikroTik (hEX or RB5009), TP-Link Omada ER605, Ubiquiti UniFi gateways, or pfSense/OPNsense on a small mini PC all work well.
  • A small UPS for the network rack: load shedding takes down your router just as surely as a cable fault does.
  • Admin access and an hour of quiet time: do this on a Saturday, not during a client demo.

Budget-wise, a capable dual-WAN router costs far less than a single lost client.

Step 1: Choose Two ISPs That Do Not Share the Same Path

This is where most offices go wrong. Two connections that ride the same upstream and the same international cable are really one connection with two bills. When SMW5 failed, both of them would slow down together.

Call each provider’s business sales team and ask three direct questions. Which upstream carrier do you buy international bandwidth from? Which submarine cable systems does your traffic normally use? What happens to my traffic when one landing station has a problem? A serious business ISP will answer. If the salesperson goes quiet, that tells you something too.

Also pay attention to the last mile. If both fibre lines enter your building through the same duct, one road dig cuts both. A mobile data connection on a different network is a cheap way to add physical diversity. Several users this week reported that certain mobile networks felt more stable than fixed lines in their area, which shows how differently providers route traffic. Yeh step skip mat karna, because everything else depends on it.

Step 2: Configure Your Router for Automatic Failover

Connect your primary ISP to WAN1 and the backup to WAN2. Then decide between two modes:

  • Failover (active/standby): all traffic uses WAN1 until it fails, then moves to WAN2. Simple and predictable, best for most offices.
  • Load balancing: traffic is spread across both links. You get more total bandwidth, but some banking portals and VPNs dislike changing source IPs mid-session.

Start with failover. On a MikroTik, you set two default routes with different distances (WAN1 distance 1, WAN2 distance 2) and enable route checking. On TP-Link Omada or UniFi, it is a toggle in the WAN settings where you mark one link as backup. On pfSense, you create a gateway group with WAN1 as Tier 1 and WAN2 as Tier 2, then point your LAN firewall rule at that group.

The detail that matters most is the health check. Many routers only check whether the cable is plugged in. During a submarine cable fault, your local link stays up while international traffic crawls. Configure the router to ping two or three external targets, such as 1.1.1.1, 8.8.8.8 and a server you control abroad, every 5 to 10 seconds. Fail over if packet loss passes around 20 per cent or latency stays above 300 ms for 30 seconds. That way the router reacts to a bad connection, not just a dead one.

Step 3: Route Critical Traffic Deliberately

Not all traffic deserves the backup link. If your 4G backup has a 50 GB cap, you do not want Windows updates and YouTube eating it in two hours.

Create policy-based routing rules so that, during failover, only critical traffic uses the backup: VoIP and video calls (Zoom, Teams, Google Meet), your VPN to client environments, Git pushes and CI/CD jobs, and your support desk. Block or throttle large updates and streaming on the backup path. Add QoS so voice gets priority. On a slow link, a call that stays clear matters more than a fast download.

While you are here, reduce how much your daily work depends on long international round trips. Use a CDN with nodes close to your users, cache package dependencies (npm, pip, Docker images) on a local mirror such as Nexus or a simple registry cache, and keep build artifacts nearby. Offices that did this were far less affected when SMW5 traffic got rerouted.

Step 4: Test Failover Before You Need It

A failover setup that has never been tested is a guess. Schedule a test once a month.

Start a video call and a large file upload, then physically unplug WAN1. Time how long it takes for the call to recover. A well-tuned setup should switch within 10 to 30 seconds; calls may freeze briefly but should not drop. Plug WAN1 back in and confirm traffic returns once the link is healthy again. Then do the harder test: leave WAN1 connected but simulate high packet loss (some routers let you do this, or you can block the health-check targets with a firewall rule) to confirm the router fails over on degradation too.

Write the results down. Note the switchover time, what broke and who noticed.

Step 5: Monitor Links and Tell Clients Early

Set up alerts so you know about a failover before your clients do. Most routers can send email or Telegram alerts, and free tools like UptimeRobot can watch your office’s public IP from outside. Keep an eye on Cloudflare Radar and PTA announcements during national incidents so you can tell a local problem from a country-wide one.

Then prepare a short client message template in advance. Something like: “Pakistan is experiencing an international cable fault. We have switched to our backup connection and all delivery dates remain on track. Live calls may have slightly lower video quality this week.” Sent early, that message protects trust. Sent after a missed meeting, it sounds like an excuse.

Common Mistakes to Avoid

Buying two lines from resellers of the same upstream. Different brand names can hide the same backbone. Ask about routes, not packages.

Forgetting power. A perfect dual-WAN setup is useless if the router and ONT switch off during load shedding. Put the router, ONTs, modem and core switch on a UPS with at least 2 hours of runtime.

Relying on “link up” checks only. During the SMW5 fault, local links stayed up while international performance dropped. Without latency and loss checks, your router will happily keep traffic on a crippled line.

Never testing. Configurations drift, passwords change and SIM data packages expire. A monthly test catches all of this.

Key Takeaways

  • Route diversity beats raw speed: two ISPs on different upstreams and cables are worth more than one very fast line.
  • Health checks must measure quality: fail over on packet loss and latency, not only when the cable is unplugged.
  • Protect the backup link: send only calls, VPNs and delivery traffic over it, with QoS for voice.
  • Test monthly and document it: measured switchover times reassure clients and catch silent failures.
  • Communicate early: a prepared client message turns an outage into proof of professionalism.

Need Expert Help?

If this feels like a lot to manage alone, TecniForge can handle the heavy lifting. Our team specializes in custom software development and AI integration. Get in touch with our experts.

Also read: SMW5 Cable Fault: 7 Lessons for Pakistani Businesses as Internet Slows Again, our earlier coverage on why this matters today.

Here is a challenge for this weekend: unplug your main internet line for five minutes and watch what happens. If your team stops working, you now know exactly what to fix before the next cable fault.


Discover more from TecniForge

Subscribe to get the latest posts sent to your email.