Alteryx Server Implementation flashcards
Alteryx Server Implementation · Build recall with a focused review session.
33 cards. Read the term and the definition here. The deck at the top of this page is optional.
- Controller role
- The Server node that hosts the Gallery web application, the AlteryxService, and the scheduler; it dispatches jobs to Workers and manages metadata.
- Worker role
- A Server node that runs the AlteryxEngine to execute workflows dispatched by the Controller; it does not host the Gallery.
- MongoDB metadata store
- The database holding Server metadata such as users, schedules, workflow versions, and results; it does not store the workflow package files.
- File store
- The shared storage location for published workflow assets and outputs, accessible to both Controller and Workers.
- Worker licensing
- The Server license determines how many Worker nodes may execute workflows; a Controller-only license cannot run jobs.
- Service account permissions
- The AlteryxService account needs Log on as a service plus read/write on the file store, log directories, and install directory.
- System Settings console
- The central configuration tool for Controller, Worker, Engine, and Gallery settings on a Server node; changes require a service restart.
- MongoDB default port
- The embedded MongoDB instance listens on port 27018 by default to avoid colliding with a standard 27017 installation.
- Engine temp directory
- The configured scratch space where the engine writes spill files; it must have free space or jobs fail.
- Gallery base URL
- The configured external URL of the Gallery, used in notifications, links, and redirects; changing it requires a service restart.
- Authentication modes
- Gallery supports built-in accounts, Integrated Windows authentication, and SAML 2.0 federation with an external identity provider.
- Curator role
- The administrative role in the Gallery that grants access to the Admin panel, users, schedules, and server-wide settings.
- Artisan role
- A Gallery role permitting a user to publish workflows and create schedules, subject to per-asset permissions.
- Viewer role
- A Gallery role permitting a user to run and consume shared workflows and results without publishing or scheduling rights.
- SAML NameID mapping
- The SAML assertion attribute matched to a Gallery user; unmatched identities are auto-provisioned or rejected per configuration.
- API access keys
- Per-user keys generated in the user profile that authenticate API calls and inherit that user's permissions.
- Publishing a workflow
- Packaging a workflow and its dependencies into the Gallery, with data connections embedded or referenced for run-time use.
- Run-as user
- The account whose credentials execute a scheduled workflow; it must have access to the required data connections and files.
- Scheduler queue
- The Controller-managed queue of pending scheduled jobs that are dispatched to available Workers for execution.
- Execution timeout
- A configured limit after which a running workflow is terminated and recorded as a failed run.
- Workflow version history
- The Gallery's retained prior published versions of a workflow, allowing an administrator to revert to an earlier version.
- Curator admin panel
- The Gallery administration interface exposing users, workflows, schedules, data connections, and server logs to Curators.
- Service and Gallery logs
- Log files under the Alteryx ProgramData directory that record job dispatch, errors, and Gallery activity for troubleshooting.
- MongoDB backup
- A metadata backup taken with mongodump against port 27018 and restored with mongorestore before starting the service.
- Run history view
- The Gallery activity view showing each workflow run's status, duration, and errors for monitoring and diagnosis.
- Log and result purging
- Scheduled cleanup of old run results and logs to prevent unbounded growth of MongoDB and the file store.
- Scale-out with Workers
- Adding Worker nodes behind one Controller to increase execution capacity, all sharing the same MongoDB metadata store.
- Worker concurrency
- The number of simultaneous jobs a Worker runs, bounded by CPU cores and memory; oversubscription causes out-of-memory failures.
- MongoDB replica set
- A multi-node MongoDB configuration providing failover for Server metadata, removing the single point of failure in a default install.
- Controller high availability
- Running a secondary Controller against shared MongoDB behind a load balancer with sticky sessions for failover.
- Queue depth monitoring
- Tracking the number of pending scheduled jobs; a sustained backlog indicates insufficient Workers or growing job runtimes.
- Run-as access failure
- A scheduled job failing with an access error usually means the run-as user's credentials or data connection permissions are invalid.
- Upgrade order
- Back up MongoDB and the file store, then upgrade Controller and Workers to the same Server version to avoid dispatch errors.