Service status & availability Crypto-only billing · No-KYC signup

Network & protection

Understand and verify IPv4 and IPv6 assignments

Read the addresses and prefixes assigned to a server, compare them with the guest configuration, and avoid locking yourself out while making changes.

Tungsto documentation · Updated · 3 min read

On this page
  1. Before you start
  2. Read the allocation correctly
  3. Inspect the running configuration
  4. Change persistent configuration safely
  5. Expected result and troubleshooting
  6. Related resources

Before you start

  • An activated server with address assignments visible in Account > Network or supplied by the operator.
  • The IPv4 and IPv6 addresses, prefix lengths, gateways, and any on-link routing instructions for this specific server.
  • SSH access plus an independent recovery path before changing persistent network configuration.
  • The operating system’s current network manager and its official configuration guide.

Read the allocation correctly

Tungsto describes an IPv4 /29 block and an IPv6 /64 allocation with each server. A prefix describes an address range; it does not by itself tell you which addresses are reserved, routed, configured as gateways, or available inside your operating system. Do not turn /29 into a claim of eight usable host addresses without the operator’s routing model.

The account’s Network page represents individual assignment records, each with an address and prefix length. It does not currently display a gateway, VLAN, routing policy, or the complete contents of a routed block. Obtain those details before making a static configuration. IPv6 commonly uses a different gateway and neighbor-discovery behavior from IPv4.

ValueWhat it tells youWhat it does not tell you
IPv4 address/prefixAn assigned IPv4 value and network sizeGateway, reserved addresses, or routing model
IPv6 address/prefixAn assigned IPv6 value and prefix sizeDefault route or whether the full /64 is routed to the host
Primary IPv4The address displayed on the server summaryAll additional addresses
Port speedThe configured link tier from the catalogueGuaranteed application throughput

Inspect the running configuration

Run read-only commands first. Interface names vary by hardware and operating system, so discover them instead of assuming eth0. Compare every active address and default route with the operator-supplied allocation. If the one-click image is already reachable, preserve the working configuration before editing it.

  • An UP interface with no expected address suggests guest configuration or provisioning is incomplete.
  • An address without a default route may permit only local-subnet traffic.
  • Multiple default routes or an unexpected metric can send replies through the wrong interface.
  • Link-local IPv6 alone does not confirm public IPv6 connectivity.
sh
ip -brief link
ip -brief address
ip route show
ip -6 route show

Change persistent configuration safely

  1. Identify whether the host uses NetworkManager, netplan, systemd-networkd, ifupdown, or another manager.
  2. Save the existing configuration and record the current addresses and routes.
  3. Confirm the exact address, prefix, and gateway with the operator; do not derive a gateway by guessing the first or last address.
  4. Arrange KVM or another independent recovery path before applying a remote network change.
  5. Use the operating system vendor’s documented method, then re-run the read-only checks from a second session.
  6. Confirm both inbound administration and outbound connectivity before closing the recovery session.

Expected result and troubleshooting

The expected result is that the running IPv4 and IPv6 configuration matches the authoritative allocation and contains the intended routes. If the Network page is empty, that is not evidence that the server has no network; it means no assignment records are available to the account. Ask the operator to populate or confirm them.

For one-family failures, compare that family’s address and default route separately. For complete loss of access after a change, stop retrying random gateway values and request operator-assisted console recovery. Include the server, region, address family, working and failing source locations, current route output if available, and the exact time of the change.