Hero mask

8 Common Peering Mistakes

(and How to Avoid Them)

Peering is often presented as a simple equation: connect to an Internet Exchange, establish a few BGP sessions, and enjoy lower latency and optimal routes.

In reality, many networks leave performance, resilience and traffic engineering opportunities on the table.

In this article, we'll look at some of the most common challenges, along with practical suggestions that can help unlock additional value over time.


1. Treating Peering as "Set and Forget"

One of the most common mistakes is assuming that Peering automatically delivers optimal routing.
Networks often establish sessions, verify that traffic is flowing, and then never revisit their routing policies again.
Months later, traffic patterns have changed, new peers have joined, and better paths might have become available.

Practical Tip

Review your peering setup at least quarterly:
- Check where your top traffic destinations
- Review newly available peers
- Verify route policies
- Evaluate latency and path changes
- Assess how much traffic is still flowing over Transit

Peering is not a one-time project. The most successful networks continuously optimise their interconnection strategy as traffic patterns evolve.

2. Ignoring PeeringDB

For many operators, PeeringDB is the first place they look before establishing a Peering relationship. Even small improvements to your profile can make it easier for potential peers to discover and connect with your network.

Practical Tip

Keep the following information current:
- NOC contacts
- Peering contacts
- Peering policy
- IXP presence
- ASN details

3. Not Using Route Servers Properly

Some networks connect to an Internet Exchange and immediately start creating dozens or even hundreds of bilateral sessions.

In many cases this creates unnecessary operational overhead.

Route Servers were specifically designed to simplify Peering by providing access to large numbers of networks through a single BGP session.

Practical Tip

Use Route Servers wherever appropriate.

Benefits typically include:
- Faster onboarding
- Reduced configuration complexity
- Easier scaling
- Faster access to new peers
- Less operational overhead

At NL-ix, members can take this a step further using the Route Server Configurator.

Instead of treating the Route Server as a black box, members can select peers based on latency, geographic region, blacklists and whitelists through a simple web interface. The resulting configuration is automatically deployed to the Route Server.

This allows networks to combine the scalability benefits of Route Server Peering with a level of control that traditionally required large numbers of bilateral sessions.

For most networks, Route Servers should be the default starting point.

If you're new to Route Servers or want to find specific information and settings for the NL-ix Route Servers, have a look at our Peering Config Guide and how to get the most value from it.

4. Assuming Every Network Uses Route Servers

Not every network though participates through Route Servers.

Some large content networks or CDNs and prefer direct bilateral peering relationships.

Practical Tip
- Review the Peering policies of your largest traffic partners.

You may discover opportunities that cannot be reached through Route Servers alone.

A healthy Peering strategy typically combines Route Server participation with selected bilateral relationships where they make operational, technical or commercial sense.

The best Peering strategies are rarely built around a single approach.

5. Choosing the Wrong Internet Exchange

Many networks evaluate an IXP based on a single metric: peer count. While peer count can be useful, it doesn't tell the whole story.

The real question is:

Can you actually reach the networks that matter to your business?

Sometimes a relatively small traffic flow can justify a dedicated Peering relationship if it supports a business-critical application.

If the networks behind these services are not present at your chosen exchange, you may find yourself relying on third-party Transit networks for some of your most critical communications.

Equally important is the structure of the exchange itself.

When evaluating an IXP, it is worth looking beyond the number of connected networks and considering HOW those networks are connected to the platform.

In a distributed exchange architecture, members can often be reached through multiple locations across the network. This creates greater flexibility in how traffic is exchanged and can provide additional routing options between participants.

The value of an exchange is therefore not only determined by who is connected, but also by how those networks can be reached.

At the end of the day, a successful Peering strategy is not just about exchanging more traffic.

It's about ensuring that the right traffic reaches the right networks through the most predictable, resilient and controllable paths available

Practical Tip

Before joining an IXP, consider:
- Which networks generate most of your traffic?
- Which networks carry your most business-critical traffic?
- Are those networks actually present on the exchange?
- Can you reach them through the Route Server?
- How much traffic can realistically be offloaded from Transit?
- What does the exchange's geographic footprint look like?

6. Forgetting Traffic Engineering

If you're operating a smaller network, you don't need to become a BGP expert overnight. However, understanding a handful of concepts such as Communities and Local Preference can already provide useful control as your network grows and can significantly improve routing decisions.

Practical Tip

Learn how to use:
- BGP Communities
- AS Path Prepending
- Selective route advertisement
- Geographic traffic steering
- Local Preference

At NL-ix, members can use BGP Communities to influence route propagation, steer traffic flows and selectively advertise routes to specific peers.

Understanding how to use Communities effectively often provides more value than adding additional peers, because it allows networks to optimise how traffic actually traverses the Internet.

Small changes in routing policy can have a surprisingly large impact on latency, resilience and transit offload.

7. Overlooking RPKI and Routing Hygiene

A surprisingly large number of routing issues can be traced back to poor route registration or missing validation mechanisms.

Incorrect routing information can affect reachability, trust and operational stability.

Practical Tip

Before expanding your Peering footprint, ensure:
- Route Objects are correct
- AS-SET information is maintained
- ROAs are configured
- RPKI validation is enabled
- Prefix filters are reviewed

Good routing hygiene should be considered a prerequisite, not an enhancement.

Before establishing new Peering sessions, it is worth reviewing your configuration against a structured configuration guide. Simple issues such as outdated route objects, missing ROAs or incorrect filters can create problems long before traffic starts flowing. You can find more on these topics here.

Many operational headaches can be prevented long before they become incidents.

8. Not Investing in BGP Knowledge

Many Peering issues are not Peering issues, but routing policy issues.

If terms such as Local Preference, MED, Communities, Route Filtering or RPKI still feel unfamiliar, investing a day in structured training can save countless hours of troubleshooting later.

Practical Tip
- Invest in BGP expertise.

NL-ix regularly hosts hands-on BGP courses covering routing fundamentals, traffic engineering, Peering over Internet Exchanges and routing security.

Even experienced engineers often discover optimisation opportunities they had previously overlooked.

The difference between a basic Peering setup and a well-engineered one is often measured in knowledge rather than hardware.

Final Thoughts

Successful Peering is not about collecting the highest number of sessions.

It's about building efficient, resilient and well-managed interconnections.

The networks that get the most value from Peering are usually not the ones with the biggest routers or the largest peer count.

They're the ones that continuously optimise, measure and refine their routing strategy.

Happy Peering!

Further Reading

If you'd like to dive deeper into the topics discussed in this article, the following resources may be useful.

Peering & Interconnection

Route Servers & Routing Policy

Training & Knowledge

Related Services