Instance network delay after IPV6 migration

Question

Hi, my instance (Ubuntu 22.04 ARM) was in default vpc and I created new VPC with IPV6 CIDR in view of recent IP charges.
To migrate my instance to new VPC, I created AMI, launched new instance in new VPC and assigned primary IPV6.    So now, `no IPv4 address or elastic IP attached to it. Just IPv6 address.`
Post this, I'm facing below issues,
1. All my websites that run on PHP started to give timeouts and respond slowly. I see many errors like below,

```
FastCGI sent in stderr: "PHP message: cURL error 28: Connection timed out after 5001 milliseconds" while reading response header from upstream, client: <>, server: www.website.com
```

2. Emails that I send from instance using AWS SES SMTP are not being sent and the scripts are just timing out.

Note: All this started to work back in BAU if I assign an elastic IP address to the instance temporarily.

Is there anything to change in /etc/hosts files or something? I've already setup Security Groups with inbound traffic for 80, 443, custom SSH port for both IPv4 & v6.
Help is appreciated in resolving this.

Accepted Answer

I've figured it out. Simple thing is now my EC2 instance is on IPv6 only and my apps are cURLing to external IPv4 sites, sending emails using AWS SES (seems not in ipv6 yet). So the timeouts. I've added elastic IP temporarily until AWS SES support IPv6.

Answer

Hi,

Is your PHP stack including curl, opensll up-to-date?

See https://kinsta.com/knowledgebase/curl-error-28/

```
Outdated software or limited server resources may cause cURL requests to time out. For example, 
outdated PHP, cURL library, or OpenSSL could trigger the cURL error 28. You can read our article 
on the ideal WordPress PHP and server requirements to see if the values match up.
```

Best,

Didier

Instance network delay after IPV6 migration

Relevant content