Installation, Configuration, and Upgrades
Installation model
Alteryx Server installs via a single installer that provisions the controller, Gallery, and (optionally) a worker on the same host. Distributed installs add remote worker nodes that must be joined to the controller using the controller's hostname and a shared secret/authentication. The System Settings interface (Alteryx System Settings) is where you configure the controller, worker, Gallery, and engine settings, including the service account, ports, and data directories.
Configuration essentials
Know that the service runs under a service account whose permissions must reach MongoDB, the data directories, and any network shares. Ports matter: the controller, Gallery (HTTP/HTTPS), MongoDB, and the message queue each use distinct ports. SSL/TLS for the Gallery is configured here. Engine settings control how many workflows a worker can run concurrently.
Upgrade traps
Upgrades require running the installer on each node in the correct order, typically controller first, then workers, then Gallery nodes, and the installer handles MongoDB schema migration. A classic trap: assuming you can upgrade a worker ahead of the controller - you cannot; version compatibility is enforced. Another trap: forgetting to back up MongoDB and the service database before upgrading, since the migration is one-way. Also know that the service must be stopped/started cleanly and that custom settings can be overwritten if not reviewed.
Worked example
You are upgrading a three-node deployment (controller+Gallery, worker A, worker B). Correct order: back up MongoDB and service DB, stop the service, run installer on the controller, then on each worker, verify versions match, restart, and smoke-test a scheduled workflow. If any node reports a version mismatch, the controller will refuse to assign work to it.
Rule of thumb
Back up first, upgrade controller before workers, and confirm every node reports the same version before declaring success.