What does the Copy report button in the Connection failed modal copy?

It copies the whole failure as plain text: the server name, the user@ip:port CtrlOps dialled, the stage it failed at (such as Dialling - TCP + SSH handshake), how long it tried before giving up, the error category and code (such as network / PreflightTimeout), and the full What happened timeline. Paste it into Discord, an email, or the CtrlOps AI Assistant instead of retyping the error. The report contains the address and the failure trace only, never your password or private key.

'Permission denied (publickey, password)'

This is the most common error. CtrlOps reached the server, but the username, password, or key was rejected. Check the username first. Different distros use different defaults: Ubuntu (AWS, DigitalOcean) uses `ubuntu`, Amazon Linux (AWS) uses `ec2-user`, Debian uses `debian`, CentOS / Rocky / AlmaLinux uses `centos` or `root`, most VPS providers use `root`. If you're using a `.pem` key, make sure permissions on your local machine are tight. From your local terminal: `chmod 600 ~/path/to/yourkey.pem`. If you're using a password, confirm password authentication is enabled on the server, SSH into it another way and check `/etc/ssh/sshd_config` for `PasswordAuthentication yes`. Click Edit on the connection card, fix the username or re-pick the key, click Connect again.

'Connection timed out' or 'Taking longer than usual, please wait...'

The server isn't responding on port 22 within 30 seconds. The cause is almost always a firewall. Check your cloud provider's security group: AWS - EC2 Console, Security Groups, Inbound rules, allow TCP 22 from your IP. DigitalOcean - Networking, Cloud Firewalls, Inbound Rules, allow SSH. GCP - VPC Network, Firewall, allow `tcp:22`. Azure - Network Security Group, Inbound security rules, allow SSH. Check the server-side firewall by SSHing in another way (cloud console, mobile hotspot, or the provider's web shell): `sudo ufw status`, `sudo ufw allow 22/tcp`, `sudo ufw reload`. If your office network blocks port 22, try a mobile hotspot or VPN to confirm. Some corporate networks block outbound SSH.

The server is reachable but the SSH service isn't accepting connections. This means SSH is stopped, crashed, or listening on a different port. Get into the server through your cloud provider's web console (the in-browser shell most providers offer), then check SSH: `sudo systemctl status ssh`, `sudo systemctl start ssh`, `sudo systemctl enable ssh`. If SSH is running on a non-standard port (some hardened servers use `2222` or `8022`), CtrlOps doesn't have a port field today, so port 22 is assumed. Either move SSH back to 22, or open a port-forward via the AI Terminal once you're in.

'Host key verification failed'

The server's identity has changed since the last time your computer talked to it. This is a security warning, not necessarily an attack. Legitimate reasons this happens: the server was rebuilt or reinstalled, the IP got reassigned to a different machine, the OS was reinstalled on the same disk. Fix on your local machine (open Terminal on Mac/Linux or PowerShell on Windows): `ssh-keygen -R `. Then try Connect again in CtrlOps. Only clear the host key if you actually expected the server to change. If the IP wasn't supposed to change, investigate before clearing.

'Network unreachable' or DNS failure

Your computer can't find a route to the server. If you're using a hostname (like `api.yoursite.com`), check DNS: `dig api.yoursite.com +short`. If the IP is wrong or empty, fix the A record at your DNS provider and wait 5 to 30 minutes for propagation. If a recent DNS change isn't sticking, try with the raw IP address as a workaround.

'Windows Server Not Supported'

CtrlOps only works with Linux servers. The supported list: Ubuntu, Debian; CentOS, RHEL, Rocky, AlmaLinux, Fedora; Amazon Linux; other mainstream Linux distributions. If you pointed at a Windows server by mistake, double-check the IP. If you genuinely need to manage Windows, that's outside CtrlOps' scope today.

The connection works, but I'm always asked for a password

You probably saved the connection without checking Remember password. Two options: For one server, open the password modal next time and tick Remember password. Switch to key-based auth: open SSH Setup from the left sidebar, under TOOLS, and use the wizard to generate a key, copy the public part to `~/.ssh/authorized_keys` on the server, then re-add the connection on the .pem Key-based Connection tab.

My `.pem` key worked yesterday but fails today

Two common causes: The server's IP changed - cloud VMs with dynamic IPs lose their address on stop/start, confirm in your provider's console, then click Edit on the connection card to update the IP. The key file moved or got new permissions - confirm the path still exists, and re-run `chmod 600 yourkey.pem`.

The connection succeeds but immediately drops

The user account on the server may have a non-interactive shell (like `/usr/sbin/nologin`), or the home directory is missing. Get in through the cloud console and check: `grep /etc/passwd`. The shell at the end of the line should be `/bin/bash` or `/bin/sh`. If it's `/bin/false` or `/usr/sbin/nologin`, fix it: `sudo usermod -s /bin/bash `.

'Too many authentication failures'

Your local SSH agent is offering too many keys before getting to the right one. Two fixes: Tell SSH to only use one specific key. Open `~/.ssh/config` on your computer and add a Host block with HostName, User, IdentityFile pointing to your key (e.g. `~/.ssh/specific_key.pem`), and `IdentitiesOnly yes`. Or, in CtrlOps, switch to the .pem Key-based Connection tab and pick the exact key file. CtrlOps then uses only that key.

SSH Connection Issues - CtrlOps Fixes

Fix SSH connection failures in CtrlOps. Diagnose timeouts, key errors, firewall rules, and security groups for AWS, GCP, Azure, DigitalOcean - in minutes.

You hit New Connection, fill in the form, click Connect SSH, and instead of landing in your server you see "Connection Failed" or "Taking longer than usual". This page walks through the most common causes and exact fixes.

How to Diagnose SSH Connection Failures

Why a "Connection failed" error appears and how to read every row of the modal.
Copying the full error report to share with support or the AI Assistant.
Fixing wrong username, wrong key, or wrong IP.
Firewall and security group fixes for AWS, GCP, Azure, DigitalOcean.
What to do when CtrlOps says "Windows Server Not Supported".
When to use the SSH Setup Wizard.

First, find the exact error

When a connection fails, CtrlOps opens a Connection failed modal that breaks the failure down rather than dumping a raw SSH string at you. Read it top to bottom:

Row	What it tells you
Server	The saved connection name, so you know which server this is about
Address	The `user@ip:port` CtrlOps actually dialled. Check this first, a typo here explains a surprising share of failures
Failed at	The stage that broke, e.g. Dialling (TCP + SSH handshake). This alone splits network problems from auth problems
Took	How long it tried before giving up. A full 45 seconds points at a firewall silently dropping packets; a failure in under a second points at auth or a refused port
Error	A `category / code` pair such as `network / PreflightTimeout`

Under that, What happened is a timestamped timeline of the attempt, marking which lines came from the app and which from SSH, so you can see exactly where it stalled. Three buttons sit at the bottom: Close, Try again, and Copy report.

The error wording still matters most. Scan the table below to find your case:

What the error says	Most likely cause
"Connection timed out"	Firewall or security group blocking port 22
"Connection refused"	SSH service isn't running on the server
"Permission denied (publickey, password)"	Wrong username, wrong password, or wrong key
"Host key verification failed"	Server's identity changed since you last connected
"Network unreachable"	DNS or routing problem
"Taking longer than usual"	Slow server, slow network, or wrong IP
"Windows Server Not Supported"	The server is running Windows; use a Linux server instead

Copy the full error report

Click Copy report (bottom right of the Connection failed modal)

This puts the whole failure on your clipboard as plain text: the server name, the address dialled, the stage it failed at, how long it took, the error category and code, and the complete What happened timeline. Paste it straight into Discord, an email, or the AI Assistant instead of retyping an error message from memory or a screenshot.

The report is safe to share. It carries the address CtrlOps dialled and the trace of what happened, never your password or the contents of your private key.

Common problems and fixes

When all else fails: gather details and ask

Before you reach out for help, collect:

Info	Where to find it
Everything about the failure itself	Click Copy report in the Connection failed modal. It captures the server, the address dialled, the stage it failed at, the timing, the error code, and the full timeline in one paste
Auth method	Password, `.pem` key, or SSH Agent
Cloud provider	AWS, GCP, DigitalOcean, etc.
Output of `ssh -vvv user@ip` from your local terminal	Run on your local machine to capture verbose SSH logs

Then bring it to:

Discord community: Join Discord
AI Assistant inside CtrlOps, paste the copied report and ask for a translation. It's surprisingly good at decoding SSH error messages.

Always keep a fallback way into your server, your cloud provider's web console, a second user, or a different SSH key on a different machine. The day your only SSH key stops working is not the day you want to be locked out.

Connection Issues