Aster Web Data / Reference

Glossary

The terms used across the Aster Web Data documentation, with a link to where each is explained.

Terms

$batch. The OData mechanism for several requests in one HTTP call. Members run in order, and an atomicity group runs in one transaction. See OData support.

$metadata. The service's description of its entity sets, types and annotations, in CSDL XML or JSON. Clients and tools read it to build their requests.

Atomicity group. Writes that succeed or fail together, sent in one $batch changeset. A set's AtomicGroup term names the group it may share with other sets. See OData support.

Commercial. A badge on a database or feature that is offered in a commercial edition. The pricing page states the terms.

Database. A named connection to a database server or file. Each DataSource is stored in one database. Clients never see the database's name. See Databases.

DataSource. A definition file, .ds.xml or .ds.yaml, that maps one entity set to a table or other source. It names the key, the fields, the navigations, and the rules that apply. See Model your data.

Entity set. A collection that a service publishes at a path, such as Orders. Each DataSource defines one set.

ETag. A value that identifies a version of a row. Clients send it in If-Match to make an update or delete conditional. A stale ETag answers 412.

Explain header. The x-aster-explain response header, which records how a request was planned: which nodes ran in a database and which in the engine. Off unless enabled. See The explain header.

Field. A property of an entity set. Its type, key, validators and computed formula come from its DataSource.

FileSource. An entity set that stores files, with actions to save, rename and remove them, and functions to list them. See AsterData vocabulary.

Hidden field. A field the caller may not read, or that a rule hides for this record. It reads as null, and the entity lists it in AsterData.Hidden.

Hierarchy. A set whose rows form a tree through a parent field. ancestors, descendants and traverse walk it. See OData support.

HTTP connection. A named entry under http: in the configuration that scripts call a REST API through, with fetch in TypeScript or http.get and the other methods in Python. It holds the base URL, the credential, the timeout, the retries and the rate limit, and scripts never see the credential. See Virtual data sources.

Navigation. A link from one entity set to another, made by a foreign key. A to-one navigation returns one entity, and a to-many navigation returns a collection.

Owned navigation. A navigation whose target entities belong to the source. They are written and deleted with it, and a deep insert or update can write them in one request.

Owner. The property that names a row's owner. When a set declares an owner, a caller sees and changes only their own rows.

Plan shape. A query's plan with its literal values removed. Metrics count requests by shape, so requests that differ only in values share one count. See Metrics and logs.

Preview. A badge that marks a database or feature offered as a preview.

Pushdown. Running part of a query in the database, rather than in the engine. The share of a read's work that is pushed down is the pushdown ratio. See What runs on your database.

Request context. Values the service declares and a client sends with each request, in the Aster-Context header or the aster-context query parameter. A rule can read them, and the service can check them against data. See Configuration.

Residual. The part of a query the engine runs in memory, after the database has run what it can. The engine's row limit, residual_rows, bounds it. See Limits.

Root. The path a service is mounted at, such as /shop. Each service has one root, and no root contains another.

Rule. A check the server applies beyond the field validators. A rule can gate a read or write, lock a field, or constrain a value. Rules are written in the DataSource or in a rules file. See Rules.

Script. A Python or TypeScript file that implements operations or a virtual data source. It reads and writes stored sets through db (aster.db in TypeScript). See Operations in code.

Seed. Rows loaded into a service's sets when the server starts. A seed is for test and demonstration data, and it is loaded once for each database.

Service. One OData API, with its own root, $metadata and namespace. A server can run several services, and they can share databases. See Configuration.

Set-based write. A PATCH or DELETE that changes many rows, sent to $each. The response reports how many rows changed.

Setup. The database's setup setting. existing uses the tables as they are. recreate drops and creates the tables, and loads the seeds. Use recreate for scratch databases only.

User source. The entity set with one row for each signed-in user. Its properties are the request facts user.<property>.

Virtual data source. An entity set served by your own code, such as a REST API, rather than by a table. See Virtual data sources.