Amazon OpenSearch Serverless

Información general

Amazon OpenSearch Serverless es una configuración serverless bajo demanda para OpenSearch, que proporciona una forma sencilla de consultar y analizar grandes volúmenes de datos. Las colecciones de OpenSearch Serverless ofrecen las mismas ventajas que los clústeres autogestionados, sin sumar configuración adicional y ajustes complejos.

Las colecciones vectoriales de búsqueda están diseñadas específicamente para búsquedas de similitudes de alto rendimiento en aplicaciones de Machine Learning (ML) e inteligencia artificial (IA) y pueden utilizarse para crear automáticamente bases de conocimiento en Bedrock.

Configuración

Instalación

Si aún no lo has hecho, configura primero la integración de Amazon Web Services.

Configuración

  1. En la página de integración de AWS, asegúrate de que OpenSearch Serverless está activado en la pestaña Metric Collection (Recopilación de métricas).
  2. Instala la integración de Datadog y Amazon OpenSearch Serverless.

Datos recopilados

Métricas

aws.aoss.2xx
(count)
The number of 2XX requests to the collection.
Shown as request
aws.aoss.3xx
(count)
The number of 3XX requests to the collection.
Shown as request
aws.aoss.4xx
(count)
The number of 4XX requests to the collection.
Shown as request
aws.aoss.5xx
(count)
The number of 5XX requests to the collection.
Shown as request
aws.aoss.active_collection
(gauge)
Indicates whether a collection is active. A value of 1 means that the collection is in an ACTIVE state. This value is emitted upon successful creation of a collection and remains 1 until you delete the collection. The metric can't have a value of 0.
aws.aoss.deleted_documents
(count)
The total number of deleted documents.
aws.aoss.indexing_ocu
(count)
The number of OpenSearch Compute Units (OCUs) used to ingest collection data. This metric applies at the account level.
aws.aoss.ingestion_data_rate
(rate)
The indexing rate in GiB per second to a collection or index. This metric only applies to bulk indexing requests.
Shown as gibibyte
aws.aoss.ingestion_document_errors
(count)
The total number of document errors during ingestion for a collection or index. After a successful bulk indexing request, writers process the request and emit errors for all failed documents within the request.
Shown as error
aws.aoss.ingestion_document_rate
(rate)
The rate per second at which documents are being ingested to a collection or index. This metric only applies to bulk indexing requests.
aws.aoss.ingestion_request_errors
(count)
The total number of bulk indexing request errors to a collection. OpenSearch Serverless emits this metric when a bulk indexing request fails for any reason, such as an authentication or availability issue.
Shown as error
aws.aoss.ingestion_request_latency
(gauge)
The latency, in seconds, for bulk write operations to a collection.
Shown as second
aws.aoss.ingestion_request_rate
(count)
The total number of bulk write operations received by a collection.
Shown as operation
aws.aoss.ingestion_request_success
(count)
The total number of successful indexing operations to a collection.
Shown as operation
aws.aoss.search_ocu
(count)
The number of OpenSearch Compute Units (OCUs) used to search collection data. This metric applies at the account level.
aws.aoss.search_request_errors
(count)
The total number of query errors for a collection.
Shown as error
aws.aoss.search_request_latency
(gauge)
The average time, in milliseconds, that it takes to complete a search operation against a collection.
Shown as millisecond
aws.aoss.search_request_rate
(rate)
The total number of search requests per minute to a collection.
aws.aoss.searchable_documents
(count)
The total number of searchable documents in a collection or index.
aws.aoss.storage_used_in_s3
(gauge)
The amount, in bytes, of Amazon S3 storage used for indexed data.
Shown as byte

Checks de servicio

Amazon OpenSearch Serverless no incluye ningún check de servicio.

Eventos

Amazon OpenSearch Serverless no incluye ningún evento.

Solucionar problemas

¿Necesitas ayuda? Ponte en contacto con el servicio de asistencia de Datadog.