BGP is the protocol that makes the Internet work by determining how packets travel between the 70,000+ autonomous systems (AS) that comprise the global network.
Why it matters
- BGP misconfigurations cause Internet outages affecting millions of users.
- Route hijacking attacks can redirect traffic through malicious networks.
- Understanding BGP is essential for network troubleshooting and security monitoring.
- Cloud and hybrid architectures increasingly rely on BGP for connectivity.
Key concepts
- Autonomous System (AS): A network under single administrative control, identified by an ASN.
- AS Path: The sequence of AS numbers a route advertisement traverses.
- Prefix announcement: An AS advertising that it can reach a specific IP range.
- Peering: Direct BGP connections between networks for traffic exchange.
- Transit: Paying another network to carry your traffic to destinations you cannot reach directly.
How BGP works
- AS establishes BGP sessions with neighbors (peers or transit providers).
- AS announces prefixes it owns or has permission to announce.
- BGP routers exchange route information and select best paths.
- Traffic flows based on the selected paths, considering AS path length, local preferences, and policies.
Security considerations
- Route hijacking: Malicious or accidental announcement of prefixes owned by others.
- RPKI (Resource Public Key Infrastructure): Cryptographically validates prefix ownership.
- ROA (Route Origin Authorization): Specifies which AS can announce a prefix.
- BGP Monitoring: Track your prefixes for unauthorized announcements.
- BGPsec: End-to-end path validation (limited deployment).
Real-world incidents
- Pakistan Telecom YouTube hijack (2008): Accidentally blocked YouTube globally.
- Cloudflare outage (2020): Configuration error caused 50% traffic loss.
- Facebook outage (2021): BGP withdrawal made Facebook unreachable for 6 hours.
Related Tools
Related Articles
View all articles
Running Local AI: The Complete Guide to Self-Hosting LLMs on Your Own Hardware
Everything you need to run large language models on hardware you own — runtimes, model formats, quantization, VRAM math, multi-GPU, Apple Silicon, and how to serve it all behind one endpoint. The hub for our local-AI series.
Read article →
How to Run an LLM Locally: A Step-by-Step Guide for Beginners
Run a large language model on your own computer in about ten minutes — no cloud, no API keys, no per-token fees. Pick a runtime, download a model, and chat privately on hardware you own.
Read article →
Local LLM Performance: What Tokens-Per-Second to Expect From Your Hardware
Why local inference is memory-bandwidth bound, what tokens/sec you'll realistically get from a 4090, a 5090, an H100, or an M-series Mac, and how model size, quantization, and context change the numbers.
Read article →
On-Prem AI for Regulated Industries: Keeping LLMs Inside Your Walls
For healthcare, finance, legal, and government, sending prompts to a third-party API is often a non-starter. Here's how to run capable AI on infrastructure you control — and meet HIPAA, data-residency, and audit requirements.
Read article →Explore More Networking
View all termsDNSSEC (DNS Security Extensions)
A suite of specifications that add cryptographic authentication to DNS responses, preventing DNS spoofing and cache poisoning attacks.
Read more →Domain Name System (DNS)
The hierarchical naming system that translates human-readable domain names into IP addresses.
Read more →IP Address Geolocation
The process of determining the geographic location of an internet-connected device using its IP address.
Read more →MAC Address
A unique hardware identifier assigned to network interfaces for local network communication.
Read more →MX Record
A DNS record type that specifies which mail servers are responsible for receiving email for a domain.
Read more →Network Interface Card (NIC)
A hardware component that connects a computer or device to a network, enabling communication through its unique MAC address.
Read more →