- Newest
- Most votes
- Most comments
This definitely sounds like a routing or ISP-level issue rather than something misconfigured on your EC2 or security groups — especially since you mentioned HTTPS (443) is wide open and you didn’t make any recent changes.
A few things to check/try:
Run traceroute or MTR to the affected IPs from different ISPs (Spectrum vs AT&T, or Miami vs Atlanta via VPN). That should help confirm if the traffic is being dropped or misrouted upstream.
Check the AWS Network Reachability Analyzer (under VPC) to rule out any internal route table or NACL issues — but honestly, since it works from some locations, it's probably not an AWS-side config problem.
Use tools like https://tools.keycdn.com/traceroute or https://www.ultratools.com/tools/ping to test external reachability from different regions.
Double-check your Elastic IP associations — if you have 4 public IPs on a single EC2, are they all correctly bound and handled at the OS/network level?
If you’re using Plesk, make sure the firewall or fail2ban didn’t auto-block certain ranges overnight. Sometimes rate-limiting or failed login attempts can silently trigger those blocks.
Definitely feels like some kind of partial route propagation issue between certain ISPs and AWS’s public IP blocks. Worth reaching out to AWS Support if this continues — they can validate route announcements from their side.
The issue you're experiencing sounds like a complex networking problem that could be caused by several factors. Given that the connectivity varies based on your location and ISP, it's likely not an issue with your EC2 instance configuration or security groups. Instead, it appears to be a routing or DNS propagation issue.
Here are some possible explanations and steps to troubleshoot:
-
DNS Propagation: If you've recently made changes to your DNS records, it might take time to propagate across different ISPs and locations. This could explain why some locations can access your sites while others can't.
-
BGP Routing Issues: There might be Border Gateway Protocol (BGP) routing problems between certain ISPs and AWS. This can cause traffic to be routed incorrectly or dropped for some paths but not others.
-
CDN or Caching: If you're using a Content Delivery Network (CDN) or caching service, it might be experiencing issues in certain regions.
-
ISP-specific Blocking: Some ISPs might be inadvertently blocking or incorrectly routing traffic to your IP addresses.
-
AWS Regional Issues: There could be ongoing AWS networking issues affecting specific regions or availability zones.
To troubleshoot:
-
Check AWS Service Health Dashboard for any reported issues in your region.
-
Verify your DNS settings and ensure they're correctly pointing to your EC2 instance's IP addresses.
-
Try accessing your sites using different DNS servers (e.g., Google's 8.8.8.8 or Cloudflare's 1.1.1.1) to rule out DNS-related issues.
-
Use traceroute or mtr to identify where the connection is failing from different locations.
-
Contact your domain registrar and AWS support to see if they can identify any issues on their end.
-
Consider using a service like Cloudflare or AWS CloudFront as a CDN, which can help mitigate routing issues.
If the problem persists, it may be necessary to engage with AWS support and potentially your ISP to investigate further. They can help identify if there are any routing or peering issues between the networks involved.
Sources
Suddenly I can't connect to my ec2 server | AWS re:Post
EC2 suddenly gets online and offline randomly | AWS re:Post
I can't connect to my EC2 instance | AWS re:Post
On some providers the DNS resolves properly but it still won't allow traffic and times out. DNS will resolve properly on some providers, and others it won't. Some providers everything works, some nothing works, some will resolve but not connect. Like I said, it acts differently from different locations.
Relevant content
- asked 5 months ago
- asked a year ago
AWS OFFICIALUpdated 10 months ago
Thank-you so much. I have been using Plesk for over a decade, and I never had this issue before. Fail2ban was blocking a bunch of IP address ranges. I removed them and turned it off and everything is working again,.