The server detail page
Everything you can do to a single server lives here, split into tabs — open any server from Servers to reach it. The page refreshes every few seconds, so status and live metrics stay current.
The header and power actions
The top of the page always shows the server's status dot, name, current state, and plan, with its public IP and region right below. Under those sits the server's handle — click it to copy — and below that, any tags you've added.
The handle: your server's permanent ID
Every server gets a handle when it's created — ekire- followed by six characters, like
ekire-7k4m2p. It's unique across all of Ekire, and it never changes.
That last part is the point. The name is yours: call a server web, rename it to api
tomorrow, call three of them staging — none of that is our business. The handle is ours, so
there's always one thing that points at exactly one server.
Quote it whenever you contact support. We can find your server from the handle alone, immediately,
with no back-and-forth about which web you mean.
Built to be read out loud
The handle deliberately leaves out the characters people mishear or mistype — no 0 or O, no 1,
i or l, no u. Whatever you read down a phone line is what we'll type in.
To the right sit the quick actions, which change with the server's state:
- Running — Console, Reboot, and Stop.
- Stopped — Start.
Before a stopped server boots, Start checks its network interfaces against the reserved IPs and private networks it holds now, and fixes any difference. If that check can't run, the start is refused with a message asking you to try again, rather than booting the server on a network setup we couldn't verify.
While a server is still provisioning, a banner lets you know it's booting; metrics and the console open as soon as it's up. If provisioning fails, the page tells you and points you to destroy and try again.
Overview
The default tab. Cards across the top show live CPU, Memory, and Disk usage as percentages, plus a Network card with current inbound (↓) and outbound (↑) rates. These update only while the server is running.
Below that:
- Specs — vCPU, memory, disk, image, region, plan, price per month, creation date, and VM ID.
- Storage — the boot disk, plus any block volumes attached to this server. Manage jumps to the Volumes page.
- Transfer — outbound data used this month against your plan's allowance, shown as a bar. Inbound traffic is always free; outbound beyond the allowance is billed per GB.
Graphs
Historical charts drawn from samples the hypervisor records. Switch between Hour, Day, and Week, and you get area charts for CPU, memory, and network (in and out), each labeled with its peak. A fresh server shows nothing here until it's been running long enough to gather a couple of samples.
Web console (VNC)
The Console tab opens an in-browser VNC session straight to the server's screen — no SSH client needed. It's available only while the server is running.
The toolbar gives you:
- Focus keyboard — hand keystrokes to the remote screen.
- Paste clipboard — types your clipboard contents into the console as keystrokes.
- Send Ctrl+Alt+Del.
- Reconnect and Fullscreen.
The connection state sits in the toolbar beside the server's name; the bar underneath shows the screen resolution and how long you've been connected.
Your way in when nothing else works
The console connects through Ekire, not over the network, so it still gets you in even when the firewall blocks everything else or you've locked yourself out of SSH. It's also how you work on a server booted into rescue mode.
Apps
The Apps tab shows every one-click app and script that ran on this server: live progress and log while it installs, then the app's address, its admin username and a Reveal credentials button. From here you can also Install an app or Run a script on a running Linux server, and Use your own domain moves an installed app to HTTPS on your domain. See One-click apps and scripts.
Networking
The Networking tab shows the Public IP and a ready-to-copy SSH command
(ssh root@your-server-ip) — each with a copy button. The Region is shown too, but isn't
copyable. If any floating IPs are attached to this server, they're listed here too, with a link to
manage them on the Reserved IPs page.
Firewall
The firewall lives at the bottom of the Networking tab. It's an On/Off toggle backed by a rules table.
- Off — all inbound traffic reaches your server.
- On — only the traffic your rules allow gets through; everything else is blocked. This covers every interface on the server: its public IP, attached reserved IPs, and private networks.
- On — not enforced — the firewall is switched on, but the host your server runs on isn't enforcing firewalls right now, so your rules aren't being applied and all inbound traffic gets through. Contact support. While that's the case, turning the firewall on or adding a rule is refused rather than shown as working. See Firewalls.
Use the Quick allow presets for the common ports — SSH (22), HTTP (80), HTTPS (443) — or add a Custom rule with a direction, action, protocol, port, source, and comment. Adding any rule turns the firewall on automatically, so a rule takes effect the moment you save it.
Turning the firewall on can lock out SSH
Once on, the firewall blocks every inbound port no rule allows — including SSH (22) unless you've added a rule for it. Allow SSH before (or as you) switch it on. Locked out anyway? The in-browser console always gets you back in.
SSH keys
Authorize a saved key to add it to this server, or Revoke one to remove it — changes push to the running server live over its guest agent, with no reboot. Ekire fully manages the server's authorized keys, so once the tab says In sync this list is exactly what's on the box. Pending means a change (including deleting a key from your keyring) hasn't reached the server yet: it's stopped, or its guest agent isn't answering. It applies on the next boot or the next background sync.
If the server is up but its keys are out of sync, the panel says so and offers Re-sync to push them now. If the server is offline there's no button — the keys go out the next time it boots. To add keys you don't have here yet, save them on the SSH Keys page first; this tab only authorizes keys already in your account.
Windows servers sign in over RDP with the Administrator password, so this tab doesn't apply to them and tells you as much.
Backups and snapshots
The Backups tab covers two kinds of restore point, plus the automated schedule.
Automated backups can be switched on or off right here; the detailed schedule (frequency and how many to keep) is configured on the Backups page. Turning them on adds a surcharge to the server's rate while enabled. On monthly billing, turning them on part-way through a billing period charges the surcharge for the rest of that period straight away; turning them off isn't refunded.
Below that, two sub-tabs separate Snapshots (point-in-time copies you take on demand) from Backups (the ones your schedule creates automatically). The first few snapshots are free; each one beyond that carries a monthly fee. A server holds at most 20 snapshots — delete one to make room.
Name your snapshot
Enter a name of 2 to 40 characters: start with a letter, then letters, numbers, hyphens, and
underscores. current is reserved and refused. A good habit is to snapshot before any risky
change, e.g. pre-upgrade.
Create it
Select Create. The new snapshot appears in the list with its timestamp.
Restore or delete later
Stop the server, then Restore to roll back to that point, or delete snapshots you no longer need.
Restoring rolls everything back
You can only restore while the server is stopped, and it stays stopped afterwards — start it yourself when you're ready. Restoring rewinds the disk to the restore point and anything changed since is lost. Your SSH keys re-sync automatically, but the root password rolls back with the disk, so if you'd changed it since, reset it afterward from Settings → Password.
The specs roll back too, so a restore taken before a resize also moves the plan and the monthly price back — and on monthly billing, a move back up charges this month's difference when the restore lands (the restore is refused if your wallet can't cover it). See Backups & snapshots. The console waits for the host to confirm the rollback before it reports back; if it can't, the restore fails and nothing is re-priced.
Your network does not roll back. The server keeps the reserved IPs and private networks it holds now: one you detached after the snapshot isn't brought back, and one you attached since stays. The firewall is left as it is now, too. If the network can't be matched right away, the restore says so and the server stays stopped. Start checks it again before booting. A server can't boot on a private network its host doesn't have, so one that's missing from the host at that moment is left off, and is taken off if the snapshot had it. It's connected again on a later start once the network is back. A network the host has but doesn't report up yet is left as it is, and the server may not start until it's up — see Private networks.
Resize
The Resize tab moves the server to a different plan — CPU and memory go up or down. Pick a target plan (the current one is marked and can't be selected), review the new monthly rate and what the resize charges now, and confirm. A custom-built server resizes onto one of the preset plans; there are no sliders here, only the catalog.
The plan list shows list prices. Once you pick a plan, the note under the list shows what this server would actually be billed after the move — after any service discount on your account (or, if a reseller set your account up, the rate your provider's account is billed for it) — next to what it's billed now, on the same basis. If that figure can't be loaded, the note shows the plan's list price and says so.
What a resize costs depends on your billing mode:
- Monthly — a billing month is paid up front, so moving to a bigger plan charges the difference the moment you confirm: the new plan's price for this billing month, less what this month has already been charged (with the backups surcharge on top of that difference if automated backups are on when the resize lands — switching them on or off while it's applying counts). The panel shows the exact amount before you confirm, and the resize is refused if your wallet can't cover it. Moving to a smaller plan charges nothing now and isn't refunded for the month already paid; the lower rate applies from your next billing month.
- Hourly — nothing is charged up front. Your wallet needs an hour of runway at the new rate, and the new rate is billed by the hour from the resize on.
Disks grow, they don't shrink
The root disk can't get smaller. If you move to a plan with a smaller disk, the server keeps its current disk and you're billed for the extra GB. A running server reboots to apply the resize; a stopped one picks up the new specs on its next start.
If you're a team admin managing someone else's account, you can only move a server to a plan that costs the same or less and includes at least as much monthly transfer. Any other plan is marked Owner only: sizing up adds to the owner's bill, and a plan with less transfer can bill the server's traffic as overage. Because each smaller catalog plan includes less transfer, sizing down is usually the owner's to do as well. See Team & collaborators.
Settings
The Settings tab is split into sub-tabs:
- General — rename the server (this is the label on your dashboard, not the machine's hostname, and never the handle) and manage its tags (up to 10, used to organize and filter your fleet).
- Password — set a new password for the
rootaccount. It's applied to the live server the moment its guest agent is up; otherwise it's saved and takes effect on the next OS rebuild. The minimum is 10 characters with at least one letter and one number, no spaces, and there's a generate button for a strong random one. It's reserved for the account owner: a team collaborator sees why instead of the form. - Rescue — boot into a small recovery Linux to repair an unbootable system or recover data. Your disk is left untouched; the server reboots to enter rescue, and you work through the web console. Exit rescue to boot back into your installed OS.
- Reinstall — lay down a fresh operating system (covered next).
We never store your password
The root password is shown to you and never kept by Ekire. Copy it somewhere safe before you leave the page — if you lose it, you'll have to reset it again.
Reinstall the OS
Under Settings → Reinstall, you can wipe the server and install a fresh operating system. The server keeps its name, plan, and public IP. It also keeps its networking: the firewall (rules, their order, and on or off), its attached reserved IPs, and its private networks with the same private IPs. All of these are set up on the fresh system before it first boots. If the reserved IPs and private networks can't be set up right then, the server boots without them rather than failing the reinstall, and they're added a moment later (the server restarts once) or, at the latest, the next time you start it. A private network that isn't up on your server's host right then is left off rather than set up (a server can't boot on a network its host doesn't have), and is connected on a later start once it's back. And if the fresh system won't boot with its reserved IPs and private networks, they're taken off and it's booted once more without them; they come back the next time you start it. Only if that second boot can't be made either does the reinstall fail. A reinstall that fails may not keep the IP — see below.
Choose an operating system
The list shows the images built on the host your server runs on — a reinstall never moves the server,
so it can only use those. Linux or Windows: choosing Windows switches the form to the RDP login model —
SSH keys don't apply, and an Administrator password is required (emailed to you, or set manually)
rather than optional. An emailed password comes with the account to sign in as: root on Linux,
Administrator on Windows.
Set login (SSH keys or password)
For a Linux image, tick the SSH keys you want authorized and, optionally, set a root password. For a Windows image there are no keys — set (or have us email) the Administrator password you'll use over RDP. A team collaborator reinstalling someone else's server can't choose the password: a new one is always emailed to the account owner, because a password the collaborator picked would outlast their access.
On Linux, password login over SSH, the forced change of an emailed password and the Ekire agent are
set up over SSH in the first few minutes after the fresh system boots. If the server's firewall is on,
the reinstall lets SSH (port 22) through it for that setup with a temporary rule at the top of your
list, and removes the rule when setup is done (normally within minutes, 30 minutes at the latest). See
Firewalls. If that temporary rule can't be added, the setup
may not get through your firewall, so the emailed password then comes with no promise of a forced
change: sign in and change it yourself (run passwd).
Type the server's name to confirm
This guard makes sure you mean it. Then select Reinstall this server.
Reinstall erases everything on the disk
A reinstall replaces the machine behind your server, so it takes the snapshots and backups with it — they live on that same machine, and a restore point can't be rolled back to across a reinstall. Copy anything you want to keep off the server first. There's no undo.
A reinstall is refused in these cases, before anything is touched:
- The server has block storage volumes. Replacing the machine would erase them — attached or detached — so the reinstall waits until you've deleted the server's volumes. Contact support if you need a reinstall that keeps them.
- The image isn't on this server's host yet. A reinstall stays on the machine your server lives on, so it can only use images built there. The picker lists only those, so you'd normally only see this if an image was withdrawn from the host while the page was open. Pick another image, or try again once it's been rolled out.
- We couldn't read the server's firewall. The firewall has to be copied before the old system is removed, so if it can't be read the reinstall doesn't start. Try again in a moment.
If a reinstall fails after the old system was removed — the new image couldn't be set up, or its disk couldn't be grown to your plan's size — the server shows Error, and its public IP is released back to the pool, where a new deploy can take it. The address isn't kept for you, so if DNS records or allow-lists point at it, plan for a new one. The server already ran and was billed for that time, so destroying it doesn't refund the month — contact support and we'll sort it out.
Destroy the server
The Destroy tab permanently removes the server. Because it can't be undone, you have to type the server's name to confirm.
While a reinstall is running, Destroy is refused: wait for the reinstall to finish (or fail), then destroy the server. If a reinstall stopped part-way and the server stays stuck in that state, it doesn't block Destroy — destroying it goes ahead. A server that is still being deployed for the first time can be destroyed straight away — that cancels the deploy.
Destroy is permanent
Destroying a server erases all of its data and disks for good — including its snapshots and backups, and any block volumes attached to it — releases its public IPv4 back to the pool, and stops billing immediately. This cannot be undone.