Troubleshooting — Networking
CIDR block refused when creating a subnet
Cause: invalid format, public range, overlap with another subnet of the VPC or with a reserved range.
Solution:
| Message | Fix |
|---|---|
| Invalid CIDR block (ex: 192.168.1.0/24) or Invalid CIDR format (e.g. 172.16.0.0/24) | Enter an IPv4 address followed by a prefix, for example 172.16.0.0/24. |
| Message stating that the range must be private | Use a range within 10.0.0.0/8, 172.16.0.0/12 or 192.168.0.0/16. |
| Message mentioning an overlap (overlaps) with an existing subnet | Choose a range that does not overlap the other subnets of the VPC (list in View Subnets). |
| Message mentioning an overlap with a reserved network | Avoid 10.244.0.0/16 and 10.96.0.0/12; prefer 172.16.0.0/12. |
VPC or subnet name refused
Cause: the name does not follow the rules, or it already exists.
Solution:
- VPC: 3 to 16 characters, lowercase letters, digits and hyphens, starting with a letter and ending with a letter or a digit. Unique within the project.
- Subnet: 1 to 63 characters, lowercase letters, digits and hyphens. Unique within the VPC.
The VPC stays in Provisioning
Cause: the platform has not finished provisioning the network.
Solution: the list updates automatically; wait a few moments. If the state does not change to Ready after a few minutes, contact support, giving the name of the VPC. If the state changes to Error, see below.
The VPC is in Error status
Cause: the platform could not apply the last change to the VPC. The VPC can keep working with its previous configuration.
Solution: when you hover over the Error badge, the console shows "The platform could not apply the last change. Contact support if this persists." Contact support, giving the name of the VPC.
The secondary interface has no address in the VM
Cause: the platform adds the interface to the VM, but the OS does not configure it automatically (this is the case on Debian and Ubuntu, which the console flags with the Manual setup required on Debian and Ubuntu warning); or the VM has not yet taken the change into account.
Solution:
- On the VM's detail page, check that VPC Networks lists the subnet and that the Internal zone contains an address of this subnet, next to the mgmt IP.
- In the VM, list the interfaces:
ip -br linkip -br addr
- If the interface exists without an address, enable DHCP on it (netplan example in Connect a VM to a VPC).
- If the interface does not appear at all, restart the VM (Restart in the Actions section).
- If the problem persists, contact support.
Two VMs on the same subnet cannot communicate
Cause: VMs on different subnets or VPCs, interface not configured in the OS, or OS firewall.
Solution:
- Compare the VPC Networks section of the two VMs: they must share the same VPC and the same subnet.
- Check in each VM that the secondary interface carries the address shown in the console (
ip -br addr). - Check the OS firewall (
sudo ufw status,sudo firewall-cmd --list-all,sudo nft list ruleset). - Test with the interface specified explicitly:
ping -c 3 -I enp2s0 <other-vm-address>
The VM lost Internet access after adding a VPC
Cause: the OS network configuration made the VPC interface the default route.
Solution: check the default route:
ip route show default
It must go through the primary interface. If it goes through the VPC interface, disable the use of DHCP routes on that interface (use-routes: false with netplan, see Connect a VM to a VPC).
Cannot delete
Cause: the VPC or the subnet is still used by VMs. The console lists the VMs concerned.
Solution: for each VM listed, open Edit > Network & Security, uncheck the VPC or the subnet, then Save. Then try the deletion again.