When you buy IP Transit, one of the first technical decisions is not only how much bandwidth you need.
It is what kind of routes you should receive from your upstream provider.
Some customers only need a default route. Others need full routes. Some networks are better served by a hybrid setup that includes both. The right choice depends on your network size, hardware, routing goals, number of upstreams, and how much control you actually need over traffic flow.
This is where many buyers overcomplicate the decision.
Full routes sound more advanced, and in some cases they are the right choice. But they are not automatically better for every IP Transit customer. A default route can be completely valid for a smaller or simpler network. At the same time, relying only on a default route can become limiting once you are multi-homed, carrying customer traffic, or trying to make smarter routing decisions between providers.
The practical question is simple:
Do you need full visibility into the global routing table, or do you just need a reliable way to reach the Internet?
What Is a Default Route in IP Transit?
A default route is the simplest way to send traffic toward the Internet.
Instead of storing a detailed route for every reachable network, your router sends traffic to a default next hop when it does not have a more specific route. In IPv4, this is usually written as 0.0.0.0/0. In IPv6, it is usually written as ::/0.
For many customers, this is enough.
If you have one upstream provider, one main edge router, and no need to compare route paths between multiple providers, a default route can keep the network simple. Your router does not need to hold the full Internet routing table. Your hardware requirements are lower. Your configuration is easier to understand. Troubleshooting is usually simpler.
A default route is common for smaller networks, enterprises, early-stage hosting environments, and customers that want Internet access without becoming deeply involved in route engineering.
The tradeoff is control.
With only a default route, your network is mostly trusting the upstream provider to make the routing decision after traffic leaves your edge. You do not see all available Internet paths from your own router. You cannot make detailed outbound decisions based on specific prefixes. You also have less visibility into how different destinations are being reached.
That is fine for some networks.
It becomes a problem when routing control starts to matter.
What Are Full Routes?
Full routes, often called a full BGP table or full Internet routing table, give your router a much more detailed view of the Internet.
Instead of receiving only a single catch-all route, your router receives a large table of routes for public Internet prefixes. This allows your network to make more specific routing decisions based on BGP attributes, prefix length, AS path, local preference, communities, and other policy choices.
This is useful when your network has multiple upstreams and you want more control over which provider is used for different destinations.
For example, if Provider A has a better path to one region and Provider B has a better path to another, full routes can help your router choose more specific outbound paths. If one upstream has route quality issues to certain networks, you have more information available to shift traffic away from that path.
Full routes are especially relevant for ISPs, WISPs, hosting providers, larger enterprises, regional networks, data centers, and infrastructure companies that need more control over routing behavior.
The tradeoff is complexity.
A full table requires routers that can handle the memory, CPU, forwarding, and convergence demands of the global routing table. It also requires someone on the team to understand BGP policy well enough to avoid creating a fragile or messy routing design.
Full routes are not just a checkbox. They are an operational responsibility.
Full Routes vs Default Route: Quick Comparison
| Routing Setup | Best For | Main Advantage | Main Tradeoff |
|---|---|---|---|
| Default route only | Simple networks, single-homed customers, smaller enterprises | Simple configuration and lower hardware requirements | Limited routing visibility and control |
| Full routes only | Networks that need detailed outbound routing decisions | More control over route selection and traffic engineering | Higher hardware and operational requirements |
| Full routes plus default route | Multi-homed networks that want control with a fallback | Better visibility with an additional safety path | Requires careful routing policy |
| Partial routes plus default route | Growing networks that want some control without a full table | Useful middle ground for selected routes or customers | Less complete visibility than full routes |
The goal is not to choose the most advanced option.
The goal is to choose the routing model that matches what your network actually needs.
When a Default Route Is Enough
A default route may be enough if your network is simple and your upstream design is straightforward.
If you only have one IP Transit provider, full routes may not give you much practical value. Your traffic has one upstream path anyway. In that case, receiving a full table may add hardware requirements and configuration complexity without changing the basic design of your network.
A default route can also make sense for enterprise networks that do not need to make detailed routing decisions. If your main requirement is stable Internet reachability and your upstream provider handles the rest, a default route may be the cleanest setup.
This is also common when the customer does not have the hardware to carry a full table. Older routers, smaller edge devices, or firewalls used as routing devices may not be suitable for full Internet routing.
In those cases, forcing full routes can create more risk than value.
When Full Routes Make Sense
Full routes become more useful when your network is multi-homed.
If you have two or more IP Transit providers, full routes give your edge routers more information to decide where traffic should go. Instead of sending everything through one preferred default path, your network can evaluate more specific destinations and apply routing policy.
This matters for operators that care about performance, resilience, and traffic engineering.
A hosting provider may want cleaner routes to major eyeball networks. A WISP may want better control over upstream paths during peak usage. A data center may need a more serious routing design for tenants. An infrastructure company may want to avoid sending too much traffic through one provider when another has a better path.
Full routes are also useful when your team wants better visibility.
If you are troubleshooting latency, path changes, reachability issues, or upstream routing behavior, having the full table gives you more context. You can see what your router is learning, which paths are being selected, and how different upstreams compare.
That visibility can be valuable.
But it only helps if someone is actually using it.
The Hybrid Option: Full Routes Plus a Default Route
Many networks do not need to think about this as an either-or decision.
A common practical approach is to receive full routes from one or more upstreams while also keeping a default route as a fallback. This can give the network detailed route visibility while still preserving a catch-all path in case a specific route is missing or policy behaves unexpectedly.
For some growing networks, partial routes plus a default route can also make sense.
A provider might send customer routes, regional routes, or selected routes while still advertising a default route for everything else. This gives the customer more control than a default-only setup without requiring the router to carry the entire global table.
The best setup depends on what you are trying to optimize.
If your goal is simplicity, default-only may be fine. If your goal is routing control, full routes become more useful. If your goal is a balance between simplicity and visibility, a hybrid model may be the right starting point.
Hardware and Operational Considerations
Before asking for full routes, make sure your edge equipment can handle them properly.
The full Internet routing table is large and continues to change over time. Your router needs enough memory, forwarding capacity, and control-plane stability to handle route updates, convergence events, and policy changes.
The risk is not just that the router accepts the table.
The risk is what happens during churn.
If the router struggles during route updates, failover events, or policy changes, full routes can create instability at the exact moment when the network needs to be stable.
This is why smaller networks should be careful about accepting full routes on underpowered equipment. A simple default route on a stable edge can be better than a full table on hardware that is barely keeping up.
Operational skill matters too.
Full routes introduce more room for policy mistakes. Local preference, AS path filtering, prefix limits, communities, and route maps can all affect traffic behavior. If those policies are not designed carefully, the network can become harder to troubleshoot.
What IP Transit Customers Should Ask Their Provider
The right questions depend on your current stage.
Before signing or upgrading an IP Transit service, ask whether the provider can support default-only, full routes, partial routes, or a hybrid setup. Ask what they recommend based on your number of upstreams, hardware, bandwidth commit, traffic profile, and growth plans.
A few questions matter more than the label:
- Can you provide full routes, default route, or both?
- Do you support BGP communities for traffic engineering or blackholing?
- What prefix limits should we configure?
- Can we start with default-only and move to full routes later?
- What route filtering do you apply on customer sessions?
These questions help you understand whether the provider is simply selling bandwidth or actually understands routing operations.
So Which One Do You Actually Need?
If you are single-homed, not doing traffic engineering, and want a simple reliable setup, a default route is probably enough.
If you are multi-homed, operating customer-facing infrastructure, or need more control over outbound routing decisions, full routes may be worth it.
If you are growing but not ready to carry the full table everywhere, a hybrid approach may be the best middle ground.
The mistake is assuming that full routes automatically mean a better network.
A better network comes from matching the routing design to the actual use case.
For many smaller customers, that means starting simple. For serious operators, it means building toward more visibility, better upstream diversity, and cleaner BGP policy over time.
How SHIFT Helps IP Transit Customers Choose the Right Setup
SHIFT works with IP Transit customers that need practical connectivity, BGP support, upstream diversity, and scalable bandwidth options.
Some customers need a simple default route. Others need full routes and more control over routing policy. Some need a phased approach that starts simple and becomes more advanced as the network grows.
The goal is not to force every customer into the same routing model.
The goal is to help each customer choose the setup that fits their network, equipment, traffic profile, and growth plan.
SHIFT can also review tailored facility opportunities for customers that want SHIFT IP Transit in locations where their infrastructure already operates. If your facility has demand, available support, and a practical deployment path, SHIFT can evaluate whether a tailored backbone extension makes sense.
Request an IP Transit Review
If you are deciding between full routes, default route, or a hybrid BGP setup, SHIFT can help you evaluate the right path.
Send your current network setup, facility location, estimated bandwidth needs, upstream requirements, and routing goals.
To request a review, email:
