Colocation Handoff Checklist for New Ubuntu Servers

What this colocation handoff checklist solves
A colocation handoff is the point where your hardware leaves the rack room and becomes a live service. On Ubuntu Server, that means you can log in safely, confirm power and network handoff details, take control of the host, and check that it is ready for applications.
This tutorial is written for a fresh Ubuntu Server 24.04 or 26.04 install on colocated hardware. It focuses on the checks that prevent the support tickets you do not want: wrong switch ports, missing out-of-band access, SSH lockouts, and servers that boot but never return after a reboot. If you are planning a move or first-time deployment, the same workflow pairs well with Hostperl dedicated server hosting and colocation-adjacent infrastructure planning.
For background reading, you may also find these useful: colocation rack units guidance, colocation power, cooling, and cross-connect planning, and Debian routing tests for broken connectivity.
Before you touch the server
Have the handoff notes ready: rack location, asset tag, switch port, power feed, out-of-band IP, and the public IPv4 address assigned to the server. For this walkthrough, the public IP example is 203.0.113.10. Replace it with the real address from your provider or remote-hands team.
Open two terminal windows on your local computer. Keep one for the first SSH session and one for testing after you create the non-root administrator account. Do not disable root access until the new login works.
Connect to the Ubuntu Server
On your local computer, start with the documented SSH entry point below.
ssh root@203.0.113.10203.0.113.10 is a documentation example. Replace it with the public IP assigned to your colocated server. If your provider uses a different initial account, use that instead, but keep the same IP example structure in your notes.
Confirm the operating system and hostname
On the VPS as root, confirm you are on Ubuntu before you change anything.
cat /etc/os-releaseYou should see Ubuntu Server 24.04 LTS or Ubuntu Server 26.04 LTS in the output. If you do not, stop here and use the matching OS instructions for that platform.
Set a clear hostname that matches the rack or service name, then reload the hostname state.
hostnamectl set-hostname server.example.comReplace server.example.com with your real host name. A clean hostname makes support handoffs, logs, and monitoring easier to read.
Create the non-root admin account
Colocation handoffs should not remain on root for routine administration. Create a named sudo user called deploy, then add your SSH key before you test it in a second session.
adduser deploySet a strong password when prompted. On a colocated machine, this still matters as a break-glass login, even if you later prefer key-based access only.
usermod -aG sudo deployThis grants sudo access on Ubuntu through the sudo group. Keep the root session open.
install -d -m 700 -o deploy -g deploy /home/deploy/.sshNow add your public key. Replace the example key with your actual public key string.
cat > /home/deploy/.ssh/authorized_keys <<'EOF'
ssh-ed25519 AAAAC3NzaC1lZDI1NTE5AAAAIExampleKeyForColocationHandoff deploy@laptop
EOFLock down ownership and permissions so SSH accepts the file.
chown deploy:deploy /home/deploy/.ssh/authorized_keys
chmod 600 /home/deploy/.ssh/authorized_keysIf you want to review the result, confirm the directory permissions now.
ls -ld /home/deploy /home/deploy/.ssh /home/deploy/.ssh/authorized_keysYou should see deploy as the owner and restrictive permissions on the SSH directory and key file.
Test the new login before changing SSH policy
On your local computer, open the second terminal and test the new account.
ssh deploy@203.0.113.10Replace 203.0.113.10 with the real server IP. If the login succeeds, confirm sudo works.
sudo -vYou should be prompted for the deploy password if your key does not already satisfy sudo. A successful response means you can manage the server without relying on root.
Update packages and install the support tools you need
Now that you have a safe admin path, bring the server fully current. On Ubuntu, use apt for updates and a small set of diagnostics that help during handoff.
sudo apt update
sudo apt -y full-upgrade
sudo apt -y install ufw chrony curl net-tools ethtool iproute2 traceroute mtr-tinyThis updates the OS, installs time sync, and adds network tools that help when you are checking switch ports, MTU, latency, and routing.
Confirm the installed kernel and package state.
uname -r
apt list --upgradableIf apt list --upgradable returns nothing relevant after the upgrade, the system is current.
Check time sync and basic hardware health
Colocated servers often sit behind different power paths and switch fabric than VPS instances. Time drift and hardware warnings are easier to catch early.
timedatectl status
systemctl status chrony --no-pagerLook for System clock synchronized: yes and an active chrony service. If the clock is wrong, log timestamps and certificate checks become unreliable.
Inspect the most recent boot messages for disk, NIC, or power-related errors.
journalctl -b -p warning --no-pagerIf you see storage reset errors or network link flaps, document them before you continue. In a colo environment, that often points to cabling, firmware, or remote-hands work.
Verify networking on the handoff port
Before you configure firewall rules, verify that the server sees the expected interface, link speed, and default route.
ip -br addr
ip route
ethtool enp1s0Replace enp1s0 with your actual interface name if Ubuntu assigned a different one. You are looking for the assigned IPv4 address, an up link, and the speed your carrier promised.
If the port is connected to the wrong switch or VLAN, ethtool often shows link up while routing still fails. In that case, compare the port handoff with the provider ticket and patch panel notes.
For a quick reachability test from the server itself, ping a known public resolver.
ping -c 3 1.1.1.1Three replies without packet loss confirm basic outbound connectivity. If this fails, move to switch, VLAN, and upstream routing checks before you harden SSH.
Set a safe UFW policy
Open SSH first, then turn on the firewall. That sequence avoids locking yourself out of a server that still needs remote access during handoff.
sudo ufw allow OpenSSH
sudo ufw allow 80/tcp
sudo ufw allow 443/tcp
sudo ufw default deny incoming
sudo ufw default allow outgoing
sudo ufw enableOpenSSH keeps your management path alive. Port 80 and 443 are optional, but they are common for a status page, reverse proxy, or application test endpoint.
Check the rules after enabling UFW.
sudo ufw status verboseYou should see the allow rules listed and the default deny policy for inbound traffic.
Harden SSH without risking a lockout
Once the new admin account works and UFW is active, tighten SSH. Do this in stages, and test after every change.
sudo cp /etc/ssh/sshd_config /etc/ssh/sshd_config.bakMake a backup first so you can roll back immediately if syntax or access breaks.
Edit the SSH daemon configuration.
sudo nano /etc/ssh/sshd_configAdd or adjust these lines inside the file:
PermitRootLogin no
PasswordAuthentication no
PubkeyAuthentication yes
AllowUsers deploySave and exit in nano with Ctrl+O, Enter, then Ctrl+X.
Test the configuration before you reload SSH.
sudo sshd -tIf the command returns no output, syntax is valid. Then reload SSH rather than restarting it, so existing sessions stay alive.
sudo systemctl reload sshFrom your second local terminal, verify that key-based login still works and that root password logins no longer do.
ssh deploy@203.0.113.10If that works, try a root login only to confirm it is blocked.
ssh root@203.0.113.10You should see a permission denied response. Keep your original root session open until this test passes.
Check the server after the handoff
At this point, the server should be reachable, secured, and ready for service deployment. Confirm the firewall, SSH daemon, and listening ports from the server side.
sudo systemctl status ssh --no-pager
sudo ss -tulpn
sudo ufw status numberedSSH should show active, and ss should list the services you expect on ports 22, 80, and 443. If extra ports appear, document them before you hand the machine over to application owners.
Also check the available disk space, since colo installs often arrive with separate data volumes.
df -h
lsblk -fIf a data disk is present but unmounted, stop and mount it before any application writes begin. A rushed handoff is how data ends up on the wrong volume.
Recovery and rollback path
If you lock yourself out, use the out-of-band console or remote-hands path from your provider. The immediate rollback steps are simple.
sudo cp /etc/ssh/sshd_config.bak /etc/ssh/sshd_config
sudo sshd -t
sudo systemctl reload sshIf UFW blocks access, disable it only long enough to restore control.
sudo ufw disableThen reapply the rules in the safe order: allow SSH first, test login, and only then re-enable the rest of your policy.
Functional smoke test
A colo handoff is not complete until you prove the server answers a real request. If you already have a simple status page or local service, test it from your local computer.
curl -I http://203.0.113.10Replace the IP with the server's actual public address. A valid HTTP response means routing, firewall policy, and the exposed service are aligned.
If you do not yet have a web service, use a host-only test by checking whether SSH still responds after a reboot.
sudo rebootWait for the machine to come back, then reconnect from your local computer.
ssh deploy@203.0.113.10Successful reconnection after reboot is the most practical persistence test for a newly handed over colo server.
Troubleshooting the most likely failures
No SSH response: From the server console, check the daemon and firewall.
sudo systemctl status ssh --no-pager
sudo ufw status verboseIf SSH is inactive, re-run sudo sshd -t and reload the service. If UFW is blocking access, allow OpenSSH first.
Wrong interface or no route: Verify the detected NIC and routing table.
ip -br link
ip route
ethtool enp1s0If the link is down or the interface name differs, match the server label against the rack handoff notes and remote-hands ticket.
SSH key rejected: Check ownership and permissions.
ls -ld /home/deploy /home/deploy/.ssh /home/deploy/.ssh/authorized_keysFix with chown deploy:deploy /home/deploy/.ssh /home/deploy/.ssh/authorized_keys and chmod 700 /home/deploy/.ssh, chmod 600 /home/deploy/.ssh/authorized_keys.
System clock wrong after boot: Check chrony and the time source.
timedatectl status
chronyc trackingIf chrony is not syncing, review outbound network access and NTP reachability.
If you are bringing a server into production or moving hardware into a colocation cabinet, Hostperl can help you choose the right deployment path and support level. For customer-facing workloads, pair this checklist with dedicated server hosting or a managed plan that fits your team’s handoff process.
Hostperl's support approach is practical: fewer surprises during migrations, clearer escalation paths for network issues, and faster recovery when a rack move or remote-hands task needs attention.
FAQ
Should I disable root login immediately after first SSH access?
No. Create and test the non-root admin account first, then disable root login after the new account is confirmed from a second terminal.
Why does colocation need more verification than a VPS?
Because you own the hardware lifecycle. Power feed, switch port, cable path, BIOS state, and disk health all matter before the server becomes a stable service.
What should I document during handoff?
Record the public IP, hostname, switch port, rack location, power path, out-of-band access details, SSH policy, and any warnings from the first boot logs.
What if the server reboots but never comes back?
Use the console or remote-hands channel to inspect boot order, RAID status, and network configuration. Then compare the live state with the handoff notes and recent changes.
For teams that run customer workloads on colocated metal, this process keeps the machine supportable from day one. It also gives you a clean path back if anything goes wrong, which is the part that matters most on a production server.
