Synapse

Use Synapse when the destination is a dedicated SQL pool you already run. Skipprd inserts into schema.table.

Before you begin

A connection string the Skipprd host can use (SQL auth or AAD, as your pool allows).
Store it in the environment.

bash

export SYNAPSE_CONNECTION_STRING="Server=tcp:....sql.azuresynapse.net,1433;..."

Configure

PythonCLIYAML

python

from skippr import Config, DataSinkSynapse, EnvRef, Pipeline

cfg = Config.discover()
syn = cfg.data_sink(
    "warehouse",
    DataSinkSynapse(
        connection_string=EnvRef("SYNAPSE_CONNECTION_STRING"),
        schema="dbo",
        table="events",
    ),
)
cfg.pipeline("events", Pipeline(data_source=cfg.get_data_source("sample"), data_sink=syn))
cfg.save()

bash

skipprd connect data-sink synapse \
  --pipeline events \
  --name warehouse \
  --connection-string '${SYNAPSE_CONNECTION_STRING}' \
  --schema dbo \
  --table events

yaml

data_sinks:
  warehouse:
    Synapse:
      connection_string: ${SYNAPSE_CONNECTION_STRING}
      schema: dbo
      table: events

Options

Key	Type	Required/Default	Description
`connection_string`	secret	Required	Synapse connection string as `${ENV}`
`schema`	string	Required	Schema
`table`	string	Required	Table

How data lands

Rows are inserted into schema.table. Retries are exactly once. Create the table first if the user cannot CREATE TABLE.

Troubleshooting

Symptom	Fix
Login failed	Refresh the connection string and export it in this shell
Firewall	Allow the Skipprd host IP on the Synapse firewall
Table missing	Create it or grant `CREATE`

Synapse ​

Before you begin ​

Configure ​

Options ​

How data lands ​

Troubleshooting ​

Next steps ​