Panel

Getting Help

When something goes wrong, the answer is almost always in the panel's logs. This page shows where to look first, and how to send the logs to the Convoy team when you can't find it yourself.

Checking the Usual Suspects

Start with the state of the containers:

Terminal
convoy ps

All five should be running, and web, worker, postgres, and redis should say healthy. Then look at whichever part isn't working:

Symptom Look at
The panel won't load, or shows an error page convoy logs web
Actions on servers never finish convoy horizon, then convoy logs worker
Usage graphs or scheduled tasks stop updating convoy logs scheduler
Certificate warnings, or HTTPS doesn't work convoy logs web, which includes the certificate requests

Sending Logs to the Convoy Team

If the logs don't make sense, send them to us with one command:

Terminal
convoy support

It collects:

  • each container's status and its logs from the last 48 hours,
  • the log of an upgrade from v4, if you ran one,
  • details about the host: operating system, CPU, memory, free disk, and Docker's version,
  • the panel's own checks: its configuration summary, the queue worker's status, pending database changes, and the error messages of the last 20 failed jobs,
  • your settings from /opt/convoy/.env, with every password, key, and token replaced by "set".

Before anything leaves the host, it removes every password, key, and token from the logs: the values in your .env, Proxmox API tokens, sign-in cookies, and credentials in URLs. Then it saves everything to a file in /opt/convoy/backups/support, shows you what it collected, and asks who may open it:

Who can open these logs?
  staff     the Convoy team only (default)
  password  anyone with the link and a password you choose
  public    anyone with the link

Press Enter to keep it to the Convoy team. Choose password or public when you want to share the logs with someone else, for example in a GitHub issue. In that case, keep in mind what secret-removal leaves in: your panel's address, the IP addresses of people who used it, your nodes' addresses, and occasionally a customer's email address.

It asks once more before uploading, then prints two links:

  • The link to share, such as https://convoypanel.com/support/CVY-7K3M-Q2XP. Paste it in Discord or your GitHub issue.
  • A private manage link. Keep it to yourself: it lets whoever has it change who can open the logs, or delete them.

Uploads are deleted after 30 days.

  1. Changing Who Can Open the Logs

    From the same host, using the code from the link:

    Terminal
    convoy support visibility CVY-7K3M-Q2XP staff
    

    To delete the upload straight away:

    Terminal
    convoy support delete CVY-7K3M-Q2XP
    

    The manage link does the same from a browser, on any machine.

  2. Reading the Logs Before They Leave

    To look first, collect without uploading:

    Terminal
    convoy support --no-upload
    

    It stops after saving the file and prints its path. Run convoy support again when you're ready to send it.

    To cover a longer stretch than 48 hours, for example a problem that started last week, pass --since:

    Terminal
    convoy support --since 7d
    
  3. If an Upgrade from v4 Failed

    Run convoy support all the same. The installer set up /opt/convoy and the convoy command before the import started, and the upload includes the import's log, which shows exactly which step failed. v4 itself isn't affected: Going Back to v4 shows how to bring it back.