Hello,
We have confirmed that ports 80 and 443 are not occupied by other programs, so there is no port conflict.
The root cause of the ingress-controller crash is that the internal cluster DNS network is unreachable — it cannot resolve Let's Encrypt domains via CoreDNS:
write udp 10.1.24.6 -> 10.43.0.10:53: write: operation not permitted
This is due to abnormal iptables/network rules, where Pod UDP traffic is blocked from reaching CoreDNS. This causes the ingress-controller to panic during Let's Encrypt initialization, resulting in all HTTP traffic failing to route.
Please try the following repair steps:
- Go to Dashboard → Servers → Restart Server — This will reset the iptables rules.
- Wait a few minutes after restarting for the ingress-controller to recover.
- If it still crashes, try Reinstalling Zeabur System components.
Additionally, please note: Server memory is tight (only 204MB available out of 7.4GB), so we recommend monitoring resource usage.
Please reply with the results.
Forum