Skip to content

ISP Network Infrastructure and WAN Links

1. What is it?

This refers to the external connectivity of a data center — how it connects to the internet, other data centers, and remote branch offices.
It is provided by Internet Service Providers (ISPs) through high-speed Wide Area Network (WAN) links.
Without this, a data center would only be a local island, not accessible to users or other sites.


2. Theoretical Definition

  • ISP Network Infrastructure → The collection of physical cables (fiber optic cables, undersea cables, satellite links), routers, and switches that deliver internet services.
  • WAN (Wide Area Network) Links → High-capacity connections (leased lines, MPLS, SD-WAN) that extend connectivity beyond the local data center to branch offices, cloud services, and end-users.

These links are designed for high availability, low latency, and redundancy.


3. Why is it important?

  • Provides access to cloud platforms (AWS, Azure, Google Cloud).
  • Enables remote employees and branch offices to connect securely.
  • Supports business continuity by ensuring always-on connectivity.
  • Allows hosting of websites, apps, and services that are publicly reachable.
  • Critical for disaster recovery (DR) sites that must synchronize data across regions.

4. How is it planned?

When planning ISP and WAN connectivity, data centers consider:

  • Redundancy → Using at least two ISPs with separate physical paths so one failure doesn’t cause downtime.
  • Bandwidth → Calculated based on expected traffic (e.g., number of users, applications, backup requirements).
  • Technology Choices:

    • Leased Lines → Dedicated fiber between the data center and ISP, guaranteed bandwidth.
    • MPLS (Multiprotocol Label Switching) → Secure private WAN service, often used by banks and enterprises.
    • SD-WAN (Software-Defined WAN) → Modern approach that uses internet connections intelligently to balance cost and performance.
    • Service Level Agreements (SLAs) → ISPs guarantee uptime, latency, and support response times.
    • Security → Firewalls, VPNs, and DDoS protection at the ISP edge.

5. Impact if not done correctly

  • Single point of failure: If only one ISP is used, downtime could last hours or days.
  • Poor user experience: Slow applications and websites due to lack of bandwidth.
  • Business losses: E-commerce, financial services, and streaming platforms could lose millions during outages.
  • Security risks: Insecure WAN links could allow man-in-the-middle attacks or data leaks.

6. Real World Example

  • Amazon Web Services (AWS) builds redundant WAN links across regions to provide low-latency global services.
  • For example, AWS India (Mumbai region) connects via undersea fiber optic cables to Singapore and Europe.
  • Enterprises like banks often use MPLS WAN links for secure financial transactions, while startups may use SD-WAN for cost savings.

7. Live Sea Cable Map

Below is the screenshot showing current network of sea cables laied world wide. Click here to dig in more

Sea Cable Map

👉 Easy Analogy:
- ISP = Highway system that connects your city (data center) to the rest of the world.
- WAN Links = The express lanes (leased lines, MPLS, SD-WAN) that guarantee faster, more reliable travel than public roads.
- Redundancy = Having two highways so even if one is blocked, traffic still flows.