Glossary
The words the app and these pages use, each with the page that says more. Where the app pairs its own word with PostgreSQL’s (a change bookmark (replication slot)), both are given.
API key
Section titled “API key”A secret a program sends to call the API as your account, read only or full. API keys.
Change bookmark (replication slot)
Section titled “Change bookmark (replication slot)”What a pipeline that reads the change log keeps on your PostgreSQL database: its place in the change log. While it exists, your database keeps every change written after that place, even while nothing reads it. You create it yourself, with the statement the pipeline shows, right before the pipeline starts. Set up a PostgreSQL source.
Change log
Section titled “Change log”The record PostgreSQL keeps of every change it writes (its write-ahead log). A pipeline reads it through logical replication, so every insert, update and delete arrives as it happens. A PostgreSQL source.
Connection test
Section titled “Connection test”The steps that check a source or destination can be used, in the order a connection is made. Each step passes, passes with a warning, or fails with a message; a step after a failed one is not run. The connection test.
Destination
Section titled “Destination”The warehouse or database a pipeline writes into: Snowflake or PostgreSQL. Snowflake, PostgreSQL.
On an API error, whose side the cause is on: customer, upstream, ours or undetermined.
The REST API.
Idempotency key
Section titled “Idempotency key”A value a program sends with each change it asks the API for, so a retry of the same request to the same path is not done twice. The REST API.
Pipeline
Section titled “Pipeline”What copies tables from a source to a destination and keeps them current: each table once, then what changes after that. A pipeline’s tables.
Proxy agent
Section titled “Proxy agent”A small program you run beside a database that has no public address; it connects out to us, so you open no inbound port. In preview and off by default: Proxy agents says what exists today.
Published tables (publication)
Section titled “Published tables (publication)”The PostgreSQL publication that names which tables’ changes your database sends to a pipeline. The pipeline shows the statements that make it. Set up a PostgreSQL source.
Reading by queries (query-based capture)
Section titled “Reading by queries (query-based capture)”A way to read a PostgreSQL source with no change log: each sync reads the rows that changed since the last. It does not see rows deleted in your database. Reading PostgreSQL by queries.
Re-sync
Section titled “Re-sync”Copying some or all of a pipeline’s tables again. Sync modes and re-sync.
Request ID
Section titled “Request ID”The identifier of one API request, in every answer’s X-Request-Id header and every error’s request_id.
The REST API.
A pipeline’s work from one start to its end, listed in the run history with its own page. History, runs and the overview.
Source
Section titled “Source”What your pipelines read from: a PostgreSQL database, or your account’s own history (the history export). A PostgreSQL source, Exporting your account’s history.
Sync mode
Section titled “Sync mode”One of four ways a table is written into your destination, chosen per table: soft delete (soft_delete), live
(live), history (history) or append-only (append). A table’s sync mode
says what each one does.
Wallet
Section titled “Wallet”Your account’s prepaid balance, which a day’s usage is charged against. Usage and billing.