Skip to content
CatalystCatalyst Docs

Panel Issues

Prerequisites

Section titled “Prerequisites”
  • Admin access to the panel, so you can open Admin → System Errors and Admin → Audit Logs.
  • Shell access on the host running the stack, to read container logs and run docker compose.

Install stopped on Fedora or RHEL (dnf5)

Section titled “Install stopped on Fedora or RHEL (dnf5)”

Fedora 41+ and RHEL/CentOS Stream 10 ship dnf5, which dropped the dnf4 config-manager --add-repo command the installer used to add Docker’s repository. An install.sh older than v1.85.0 aborts at that step, so Docker is never installed and the stack is never created.

  1. Re-download the current installer (v1.85.0 or later) and run it again. It detects dnf5, installs dnf5-plugins and uses dnf config-manager addrepo --from-repofile=; on dnf4 it keeps using dnf-plugins-core and --add-repo.
  2. To repair an old script by hand: sudo dnf install -y dnf5-plugins and sudo dnf config-manager addrepo --from-repofile=https://download.docker.com/linux/fedora/docker-ce.repo (use the centos path on RHEL/CentOS), then re-run the installer.

Panel will not start after an update

Section titled “Panel will not start after an update”
  1. Read the backend container logs first (docker compose logs backend from the install directory). Most failures announce themselves: a migration waiting on a lock, a renamed environment variable (check the release notes), or an unreachable database.
  2. Confirm the PostgreSQL and Redis containers are up (docker compose ps) before blaming the backend.
  3. Roll back per updating Catalyst if the cause is not obvious within minutes — a down panel blocks all management.

HTTP 500s in the UI

Section titled “HTTP 500s in the UI”
  • One page broken, the rest fine: open Admin → System Errors and filter by Component to see the persisted exception, then check backend logs at that timestamp. A missing column after a partial migration looks exactly like this — never interrupt migrations.
  • Everything broken: database or Redis connectivity. Verify credentials, connectivity, and free disk on the database host.

Nobody can log in

Section titled “Nobody can log in”
  1. Can you reach the panel URL at all? If not, this is networking, not auth.
  2. Wrong password for everyone including admins: check the backend logs for auth/database errors rather than resetting passwords blindly. Sign-in failures are recorded in Admin → Audit Logs as login_failed.
  3. Invites or resets never arrive: mail configuration (configuration), spam folders, then mail provider logs.

Pages load but data is stale

Section titled “Pages load but data is stale”

The UI caches aggressively. Hard-refresh, then check whether realtime streams are connecting (realtime events) — a broken event stream looks like frozen data while REST keeps working.

Collect for a bug report

Section titled “Collect for a bug report”

Panel version, browser and version, exact steps, backend log lines for the incident window, and whether the problem survives a private window (rules out extensions and cookies). Admin → System Errors → Download diagnostics captures panel, node, and server logs in one redacted ZIP — see diagnostics bundle.

Section titled “Related”