The Challenge: Cutting Connectivity Costs Without Losing Resilience
The client needed a network failover testing outcome it could trust, not just a lower invoice.
When a financial services firm asks its IT provider to bring connectivity costs down, the easy answer is usually the wrong one. Strip out redundancy, consolidate onto a single circuit, and the invoice shrinks along with the safety margin. For a firm managing client assets, that trade-off isn’t acceptable, and it wasn’t what the client was asking for either. They wanted a lower bill without giving up anything on reliability, full stop.
Their existing setup relied in part on an off-net circuit from a third-party vendor, feeding into their infrastructure at Teraco. It worked, but it came with a familiar frustration for anyone who has dealt with a vendor that doesn’t own its own last mile. Any change, from a service upgrade to a new connection type, meant paperwork, planning sessions, and a timeline measured in months rather than days. The circuit itself wasn’t the problem. The rigidity built around it was, and that rigidity was quietly limiting what the client could ask for going forward.
The Approach: Designing for Flexibility, Not Just Cost
Rather than treating this as a straightforward like-for-like replacement, our network team started by mapping out what was already coming into the client’s environment before proposing anything new. Adding another physical circuit was the obvious option, and also the wrong one. More infrastructure means more to support, more complexity for the next engineer who has to troubleshoot it under pressure, and more points where something can go wrong later, none of which serves a client who came to us asking for less overhead, not more.
Instead, the team identified an existing on-net circuit, already terminating in the right place, that could take over the role the existing provider had been playing. Because Si Futures owns and operates that last mile directly, taking on this piece of the client’s connectivity meant we could offer a level of flexibility an off-net reseller simply can’t match. Where a vendor might need weeks of internal sign-off to provision a Layer 2 service into a hosting environment, owning the infrastructure means that same conversation can start and finish in a single day. That difference shaped the whole design, not just the pricing.
The Solution: A Migration Path Built for Network Failover Testing
The migration itself followed a structured, low-drama path designed to avoid any disruption to a business that genuinely cannot afford downtime:
- Decommissioned the existing off-net circuit and brought the client onto an existing Si Futures on-net circuit via one of our carrier relationships, avoiding the need for any new physical infrastructure
- Configured BGP on the new link, including conditional advertisement, so that a failure on the primary path would trigger an automatic failover to the secondary without any manual intervention
- Coordinated the cutover directly with the carrier and the client, keeping the number of parties involved to a minimum so scheduling didn’t become its own project
- Ran a full planned failover test once the new circuit was live and stable, deliberately taking the primary link down to simulate a hard failure and confirm the BGP configuration behaved exactly as designed under real conditions
- Cleaned up all legacy configuration left over from the migration itself, removing anything no longer required so the environment stays simple and clear for whoever has to work on it next
The failover test wasn’t a formality. It was the point where design meets reality, and it’s where a lot of “resilient” networks quietly reveal gaps that only show up under an actual fault condition. This one didn’t, and the carrier coordination that made the cutover possible turned out to be far more straightforward than the team had braced for.
The Outcome: What Network Failover Testing Proved
The failover completed in four seconds. From the moment the primary link was deliberately taken down to the moment traffic was flowing cleanly over the secondary path, four seconds passed, with no manual intervention and no surprises. The entire test, including preparation and verification, took roughly half an hour from start to finish.
That result matters more than the number itself suggests. It confirms that the cost reduction the client asked for didn’t come at the expense of the resilience they were relying on every day. The environment behaves exactly as it did before the change, just at a lower monthly cost and with a provider who can now offer considerably more flexibility around whatever comes next. Nothing about the client’s day-to-day experience of their network changed, which was precisely the point.
What Made the Difference: Owning the Last Mile
The decision that shaped this project happened before any configuration work began: choosing to route the client onto infrastructure Si Futures already owned rather than defaulting to a new build. It would have been easier, in some ways, to simply lay a new circuit and call the job done. It would also have added complexity the client didn’t need and support overhead the team would have been carrying indefinitely, for no real benefit to the person paying the invoice.
Using infrastructure Si Futures already owned and operated meant the project stayed genuinely simple, for the client and for the engineers who will maintain it long after this migration is forgotten. It also meant the flexibility advantage wasn’t theoretical. Owning the last mile is what turns “we can look into that” into “we can have that live this week,” and that’s the kind of difference a client only notices the next time they need something changed quickly.
Client Impact: Lower Cost, Same Reliability
For a business where reducing IT spend can feel like a risk in itself, the outcome here was a genuine balancing act pulled off cleanly: lower cost, same reliability, nothing lost. The client’s response has been straightforward relief. They’re satisfied that the change delivered exactly what it promised, without any of the usual anxiety that comes with touching core connectivity for a business that depends on it every trading day.
Longer term, the real value sits in what this unlocks rather than what it saved. With Si Futures now owning the circuit end to end, the client has a direct path to faster, simpler changes going forward, whether that’s additional services, new connections into hosting environments, or adjustments as their infrastructure needs evolve over time.
