Q: We have a BGP VPN connection established over IPSec from our data center to AWS. The tunnel is "UP", but large file transfers keep freezing or failing halfway through, while small pings and SSH commands work fine. What is happening?
This is absolutely an MTU (Maximum Transmission Unit) mismatch issue.
Want to master this scenario in a live sandbox? KodeKloud's Istio Service Mesh & Advanced Kubernetes Networking Course covers this exact problem with hands-on terminal drills.
🛠️ Production Runbook & Step-by-Step Resolution
Direct Answer: "In our multi-VPC setup, services in private subnets ran into this exact routing obstacle. The interviewer is testing: MTU (Maximum Transmission Unit), MSS, Path MTU Discovery, IPsec header overhead.. I structure my answer around systematic triage first, root cause analysis second, and permanent remediation third."
Initial Diagnostics & Root Cause Analysis
This is absolutely an MTU (Maximum Transmission Unit) mismatch issue.
- Enable
TCP MSS Clampingon the VPN router (modifying the TCP handshake to force a lower MSS, e.g., 1350). - Allow ICMP type 3 code 4 (Fragmentation Needed) on all firewalls.
- Lower the MTU manually on the host interfaces.
Remediation & Permanent Safeguards
Standard ethernet MTU is 1500 bytes. However, IPsec VPN tunnels add encryption headers (ESP, tunnel mode) which consume ~50-80 bytes. If an application tries to send a full 1500-byte packet with the "Don't Fragment" (DF) bit set, the VPN router drops it because it exceeds the tunnel's inner MTU, and sends an ICMP "Fragmentation Needed" message back. However, if firewalls along the path are blocking these ICMP messages (breaking Path MTU Discovery), the application never slows down its packet size. Known as a "PMTUD Blackhole". To fix:
- Enable TCP MSS Clamping on the VPN router (modifying the TCP handshake to force a lower MSS, e.g....
- Allow ICMP type 3 code 4 (Fragmentation Needed) on all firewalls.
- Lower the MTU manually on the host interfaces.
Senior hiring managers and bar raisers evaluate how you prioritize during outages. Avoid these common anti-patterns:
- Treating stateless Network ACLs as stateful firewalls without specifying outbound ephemeral return ports when configuring Networking connection established Security Groups track connection state automatically, whereas stateless Network ACLs require explicit ephemeral port rules (1024-65535) for return traffic.
- Opening short-lived non-pooled TCP client connections, exhausting ephemeral ports in TIME_WAIT state when configuring Networking connection established Rapidly opening and closing short-lived TCP sockets without connection pooling exhausts ephemeral ports, causing connection refusal errors.
- Directing load balancer health checks to heavy computational endpoints rather than dedicated health routes when configuring Networking connection established Deep health check paths amplify database load during traffic spikes and risk cascading eviction of healthy application nodes.
- Triggering BGP route dampening by rapidly announcing and withdrawing multihomed network prefixes when configuring Networking connection established Rapidly flapping BGP route announcements triggers upstream ISP route dampening, effectively blackholing network traffic for hours.