Direct Connect
Last updated
Was this helpful?
Last updated
Was this helpful?
AWS Direct Connect is a dedicated network connection service that creates a private, high-bandwidth link between your on-premises infrastructure and AWS. Think of it as a private highway between your data center and Amazon's cloud, bypassing the public internet.
The connection if private.
Compared to internet based connections, it reduce network costs, increase bandwidth throughput, and provide more consistent network experience.
There are two types of Direct Connect Connection:
Dedicated Connection: A Physical Ethernet connection associated with a dedicated connection through the AWS Direct Connect console, the CLI, or the API.
Hosted Connection: A physical Ethernet connection that an AWS Direct Connect Partner provision on behalf of a customer.
Key benefits include:
Performance
Consistent, predictable network performance
Lower latency compared to internet-based connections
Higher bandwidth options (1Gbps, 10Gbps, 100Gbps, 400Gbps)
Security
Private connection that doesn't traverse the public internet
Data travels through dedicated fibre optic lines
Compatible with all AWS security features
Cost Savings
Reduced data transfer costs compared to internet-based transfer
More predictable network costs
Better economics for high-volume data transfer
Connection Options
Dedicated Connection: Direct 1Gbps or 10Gbps connection
Hosted Connection: Smaller capacities through AWS partners
LAG (Link Aggregation Groups): Multiple connections bundled together
Common use cases:
Hybrid cloud architectures
Large-scale data migration
Real-time data processing
Disaster recovery
Business-critical applications requiring consistent performance
AWS Direct Connect's availability depends on how you configure your connections. Here's a breakdown:
Single Connection (99.9% availability):
One Direct Connect connection
One customer router
One AWS Direct Connect location
Vulnerable to single points of failure
Active/Active Configuration (99.99% availability):
Two or more Direct Connect connections
Multiple customer routers
Different AWS Direct Connect locations
Both connections actively used
Traffic distributed across connections using BGP
Automatic failover if one connection fails
Active/Standby Configuration:
Similar to Active/Active but with one primary and one backup connection
Secondary connection only used if primary fails
Slightly higher latency during failover compared to Active/Active
Best Practices for High Availability:
Use multiple Direct Connect locations
Set up connections from different service providers
Implement redundant customer routers
Configure BGP routing for automatic failover
Consider using Direct Connect Gateway for multi-region connectivity
AWS recommends implementing redundant Direct Connect connections through different Direct Connect locations for business-critical applications to achieve the highest availability.
Additionally, you can maintain a backup VPN connection as a last resort failover option, though this would route through the public internet and have different performance characteristics.
VPNs allow private communication, but it still traverses the public internet to get the data delivered. While secure, it can be slow.
Direct Connect is:
Fast
Secure
Reliable
Able to take massive throughput