Get started with Elglide

• Direct — pick it when the source is already reachable from outside: a public REST API, an Amazon S3 bucket, an internet-facing SFTP server.
• SSH tunnel — pick it when your database sits on a private network and you already run, or can run, a bastion.
• Hybrid agent — the only option for a network file share (SMB) or a folder on a server, and the one that keeps your data off our cloud.

1. Open Connections, stay on the Sources tab and press Add (or + on the Sources panel of the canvas).
2. Pick the source in Connector Category — a database, a file location such as SFTP, a network share (SMB) or Amazon S3, or REST API.
3. Fill in the host and sign-in details, and set Connectivity to the method from step 1.
4. Press Test Connection. For a database, tick the databases to sync — each becomes its own connection — then press Create Connection.
5. For files, add one or more File Sets (a folder and a file-name pattern); for a REST API, add Endpoints and press Detect Schema.

Done when the connection is listed under Sources and Test Connection passes.

1. Open Connections, switch to the Destinations tab and press Add (or + on the Destinations panel of the canvas).
2. Enter the Account, choose the authentication Method (Key Pair or OAuth Token) and enter the Username and key.
3. Set the default Role and Warehouse.
4. Press Test Connection, tick the Snowflake database to load into, and press Create Connection.

Done when the connection is listed under Destinations and Test Connection passes.

• Canvas (the Overview page): drag the source card onto the Snowflake card, or drag from the source's port to the destination.
• List: on Connections, open the source's ⋯ menu, choose Wire new destination, then pick Snowflake in Select destination….

Either way the pipeline screen opens with the source's tables, views, file sets or endpoints. The pipeline is saved when you save its first tasks in step 6.

1. Set the filter to Show tables/views without a task (or Show all).
2. Find objects with Search; they are grouped as Tables, Views, File sets or Endpoints.
3. Tick the ones to sync and press Save.
4. Press Sync on a task, or wait for its schedule.

Done when the first run shows as succeeded in the task’s Sync History and the table is in Snowflake.

• Schema Map — the Snowflake schema each source schema loads into. It starts as the source schema’s name in upper case.
• Default Schedule — Manual only, minutely, hourly, daily, weekly, monthly or a custom cron, in your time zone. A task without its own schedule follows it.
• Soft delete — optional, database sources only. Decide before the first run, because the option locks once a task is triggered. See Soft delete.

Done when you press Save and the object list shows a Next Run for your tasks.

• Schema → Schema Map & Column Filter — choose which columns load and review their Snowflake types.
• Schema → Row Filter — a WHERE condition that limits which source rows load. Press Validate before Save.
• Settings → Sync Schedule — a schedule for this task alone, overriding the pipeline default.
• Settings → Sync Mode — for databases, the Full → RowHash → Watermark ladder or change data capture; for files and APIs, Full or Incremental with Replace, Upsert or Append.

Done when the next run in Sync History reflects your change.

• Data Quality → Audit Rules — compare Snowflake with the source after each run; press Audit Now to check at once. Database sources only.
• Data Quality → On Audit Failure — retry automatically, fall back to a full sync, and email when a mismatch is final.
• Schema → Schema Evolution — what to do when a source column is added, dropped or changes type: follow the change, or stop and notify.