On paper, IP transit looks simple: you pay someone to carry your traffic to the rest of the Internet. In reality, the difference between a solid transit provider and a bad one is the difference between “it just works” and “we’re chasing ghosts at 2 a.m.” Price per Mbps only tells a tiny part of the story; the rest hides in routing behavior, capacity planning, and how seriously your provider treats the health of their network.
This guide walks through the biggest IP transit red flags to watch for when you’re choosing (or re‑evaluating) an upstream. If a few of these sound uncomfortably familiar, it might be time to rethink who carries your bits.
When Your Transit Provider Chose You (Not the Other Way Around)
A surprisingly common pattern: your “IP transit provider” is simply whoever your data center sales rep recommended when you signed the contract. You never evaluated them directly; they were bundled into the checklist.
That’s a red flag because:
- You probably don’t know where their edge router actually is (same building vs extended from somewhere else).
- You haven’t looked at how they perform to the destinations your customers care about.
- You may be missing better options that are already cross‑connected in the same facility.
If you can’t answer “who is our upstream, where is their edge, and why did we choose them?”, you didn’t really choose them.
Single‑Homed and Proud of It
Running an access network, hosting platform, or SaaS edge on a single transit provider is betting your uptime, latency, and reputation on one company’s design and luck.
The red flag isn’t just “we only have one upstream.” It’s when that fact is justified as a long‑term strategy, not a temporary starting point. A single‑homed network:
- Inherits every congestion event and outage from that provider.
- Has no way to route around bad paths or failures.
- Can’t use BGP policies to influence how traffic enters or leaves.
At small scale, single‑homing is understandable. At real, paying‑customer scale, it’s a sign that resilience hasn’t been taken seriously yet.
“Cheapest Per Mbps” as the Only Metric
Price per Mbps matters; nobody wants to overpay. But “we bought the cheapest bandwidth we could find” is usually followed, months later, by:
- Mysterious packet loss and jitter during peak hours.
- Strange detours through far‑away regions.
- Support tickets from gamers, VoIP users, and SaaS customers.
Cheap transit can work well if it’s engineered properly and used in the right way, but if cost is the only decision factor, you’re likely paying a hidden tax in performance and support pain.
A healthier mindset is: “We want good value per Mbps from a provider that cares about paths, capacity, and reputation,” not “whatever is cheapest on the price sheet.”
No Transparency About Capacity and Topology
You plug into a 10G or 100G port, but what’s behind it?
A classic red flag is a provider that sells lots of edge capacity without a clear story about their backbone capacity and how they engineer it. Signs to watch for:
- Very high port densities in a rack, but vague answers about upstream links.
- Repeated peak‑time slowdowns that magically “resolve themselves.”
- No documentation or discussion about how they plan for growth.
If you ask basic questions like “What does our traffic path look like from this DC to major eyeball ISPs or clouds?” and you get hand‑wavy answers, assume you’re attaching to an oversubscribed or poorly planned core.
Weak or Non‑Existent Routing Tools
Modern IP transit isn’t just “we give you the full table.” It should also give you knobs and levers to shape traffic.
Red flags here include:
- No documented BGP communities to control local preference, prepend behavior, or blackholing.
- No clear guidance on how to influence inbound or outbound routing.
- “We just do default BGP, you don’t need to worry about it.”
Without these tools, you can’t:
- Prefer one path over another for certain destinations.
- Drain traffic quickly during maintenance or partial outages.
- Offer smarter routing to your own customers.
If you’re serious enough to care about transit quality, you’re serious enough to need more than a single default route.
Route Hygiene and Security Are an Afterthought
The global routing system is fragile enough without careless players. Another major red flag is a provider that treats route security and data hygiene as optional.
Warning signs:
- No talk of route filtering, RPKI validation, or proper registry data.
- Encouraging customers to announce anything without checks.
- A history of leaks, hijacks, or sloppy announcements.
When your upstream is careless here, you share the fallout. Your prefixes may be filtered more aggressively; your reachability can fluctuate; your own reputation suffers by association.
Poor Operational Culture and “It Looks Fine Here” Support
Even good networks hit incidents. What matters is how they respond.
Support red flags include:
- Every ticket gets the same answer: “Everything looks fine on our side.”
- No willingness to share traceroutes, graphs, or a real explanation.
- Long delays before anyone technically competent looks at your case.
These behaviors usually reflect deeper issues: weak monitoring, lack of on‑call expertise, or a culture that isn’t oriented around uptime and customer impact.
A healthy provider doesn’t have to be perfect, but they should be able to say what broke, why, and what they’re doing to prevent it next time.
Putting It Together – A Simple Red Flag Table
Here’s a quick way to see how many boxes your current transit ticks.
| Area | Red flag behavior | What you want instead |
|---|---|---|
| How you chose them | “DC sales guy recommended them” | Deliberate evaluation of paths and reputation |
| Redundancy | Single upstream, long‑term | At least two upstreams and a plan to add more |
| Pricing mindset | Only “cheapest Mbps” matters | Balance of price, performance, and reputation |
| Capacity transparency | Vague about backbone, frequent peak issues | Clear design, honest capacity discussion |
| Routing control | No useful BGP communities or docs | Documented tools for traffic engineering |
| Route security | Little interest in filtering/RPKI/clean data | Visible commitment to route hygiene |
| Support culture | “Looks fine here” with no detail | Real diagnostics, open comms during incidents |
If you see yourself on the left side of this table more often than the right, it’s probably time to review your upstream choices.
Choosing Better Transit (Without Going Overboard)
Fixing IP transit doesn’t have to mean a massive overhaul overnight. Often, the most realistic steps are:
- Add a second, higher‑quality upstream alongside your existing one.
- Start using the routing tools your providers offer (communities, local‑pref, prepending).
- Gradually move more critical or latency‑sensitive traffic toward the better paths.
- Over time, retire or downgrade providers that consistently underperform.
If your network is big enough, plugging into one or more IXPs can also shift a lot of traffic onto shorter, more predictable routes, reducing the load on transit links altogether.
Getting a Second Opinion on Your Upstreams
If reading through these IP transit red flags makes you uneasy, that’s actually a good place to be, it means you’re seeing the risk before the next big incident forces the issue.
If you’d like a second opinion on your current upstream mix, how they’re routing your traffic, where congestion might be hiding, and what a cleaner design could look like, reach out to sales@shifthosting.com with a brief description of your network and typical traffic patterns. A quick review of real paths and a few key destinations is often enough to spot the easiest wins.
