OpenSearch

Use the @bijection/connectors/opensearch entrypoint to declare OpenSearch resources.

Availability

This page describes the pinned source package. Published availability is unestablished for this artifact. Source build and package availability.

Resources and modes

Resource	Mode	Description
Index	`scan`	scan

Declare a resource in application code, then bind deployment-specific settings and credentials. See typed resources, connections and source inspection.

Index

scan: scan

The following example matches configuration standard.

import { opensearch } from "@bijection/connectors/opensearch";
export const source = opensearch.index.scan({ deletions: { everyHours: 24 } });

Inferred record types

{
  "key": string;
  "document": any;
}

Configuration: standard

Case: opensearch/index/scan/standard. Connection and resource configuration are declared by this exact case. Credential kind: opensearch. Keep credential material in the private credential inventory.

Setting	Type	Required	Owner
`endpoint`	string	Yes	Shared connection
`index`	string	Yes	Resource binding

Complete configuration contract

{
  "setup": [],
  "resource": {
    "access": {},
    "defaults": {},
    "connector": "opensearch",
    "output": {
      "kind": "single",
      "collection": "records"
    },
    "credential_kind": "opensearch",
    "connection_fields": [
      "endpoint"
    ]
  },
  "connection_setup": {
    "type": "object",
    "value": {
      "endpoint": {
        "fieldType": {
          "type": "string"
        },
        "optional": false
      },
      "index": {
        "fieldType": {
          "type": "string"
        },
        "optional": false
      }
    }
  },
  "execution": null
}

Native admission analysis is available for this exact declaration. Collection: records

Contract	Meaning
Initial	Pinned snapshot
Changes	Repeated observations of current records
Deletions	Reconciliation
Resume	Opaque cursor
Gap	Explicit repair required after a gap
Retention	Unspecified

Complete native profile

{
  "movement": {
    "exchanges": {},
    "nativeInputLimitBytes": "0",
    "steps": [
      {
        "id": "acquire.1",
        "inputs": [],
        "action": {
          "operation": "acquire",
          "subject": "native input",
          "limitBytes": "0"
        }
      },
      {
        "id": "retain.1",
        "inputs": [
          "acquire.1"
        ],
        "action": {
          "operation": "retain",
          "subject": "acquisition evidence",
          "reason": "recovery",
          "release": "acquisition_custody"
        }
      },
      {
        "id": "interpret.1",
        "inputs": [
          "retain.1"
        ],
        "action": {
          "operation": "interpret",
          "subject": "acquired input",
          "produces": "records"
        }
      },
      {
        "id": "materialize.1",
        "inputs": [
          "interpret.1"
        ],
        "action": {
          "operation": "materialize",
          "subject": "source rows",
          "reason": "serving",
          "release": "replacement"
        }
      },
      {
        "id": "publish.1",
        "inputs": [
          "materialize.1"
        ],
        "action": {
          "operation": "publish",
          "subject": "source rows",
          "visibility": "with_acquisition"
        }
      }
    ]
  },
  "initial": "pinned_snapshot",
  "changes": "observations",
  "deletions": "reconciliation",
  "resume": "opaque_cursor",
  "retentionMs": null,
  "gap": "explicit_repair",
  "limits": {
    "deletion_sweep_hours": "24",
    "exchange_evidence_bytes": "524288",
    "publication_page_records": "128",
    "record_fields": "32"
  },
  "restrictions": [
    "Repeated point-in-time scans in shard document order; an index keeps no deletion tombstones, so no incremental read is claimed",
    "Deletions are published when a key lookup answers absent; a sweep runs at most every deletion_sweep_hours"
  ],
  "commands": {}
}

Commands, occurrences and artifacts

{
  "records": {
    "records": {
      "protocol": "record_observation",
      "coverage": "records",
      "consistency": "observations",
      "atomicity": "record",
      "history": "observations",
      "recordRefresh": false,
      "reconciliation": false,
      "notifications": false,
      "snapshotMaxRows": null,
      "profile": {
        "movement": {
          "exchanges": {},
          "nativeInputLimitBytes": "0",
          "steps": [
            {
              "id": "acquire.1",
              "inputs": [],
              "action": {
                "operation": "acquire",
                "subject": "native input",
                "limitBytes": "0"
              }
            },
            {
              "id": "retain.1",
              "inputs": [
                "acquire.1"
              ],
              "action": {
                "operation": "retain",
                "subject": "acquisition evidence",
                "reason": "recovery",
                "release": "acquisition_custody"
              }
            },
            {
              "id": "interpret.1",
              "inputs": [
                "retain.1"
              ],
              "action": {
                "operation": "interpret",
                "subject": "acquired input",
                "produces": "records"
              }
            },
            {
              "id": "materialize.1",
              "inputs": [
                "interpret.1"
              ],
              "action": {
                "operation": "materialize",
                "subject": "source rows",
                "reason": "serving",
                "release": "replacement"
              }
            },
            {
              "id": "publish.1",
              "inputs": [
                "materialize.1"
              ],
              "action": {
                "operation": "publish",
                "subject": "source rows",
                "visibility": "with_acquisition"
              }
            }
          ]
        },
        "initial": "pinned_snapshot",
        "changes": "observations",
        "deletions": "reconciliation",
        "resume": "opaque_cursor",
        "retentionMs": null,
        "gap": "explicit_repair",
        "limits": {
          "deletion_sweep_hours": "24",
          "exchange_evidence_bytes": "524288",
          "publication_page_records": "128",
          "record_fields": "32"
        },
        "restrictions": [
          "Repeated point-in-time scans in shard document order; an index keeps no deletion tombstones, so no incremental read is claimed",
          "Deletions are published when a key lookup answers absent; a sweep runs at most every deletion_sweep_hours"
        ],
        "commands": {}
      }
    }
  },
  "occurrences": null,
  "artifacts": {
    "raw_archive": false,
    "native_file_transfer_limit_bytes": null,
    "external_versions": false,
    "files": {}
  },
  "commands": {}
}

Recorded checks

Requirement	Scope	Evidence	Result
page-publication-recovery	{“kind”:“collection”,“name”:“records”}	The native store reader exists in platform source with unit tests of its wire decoding; no provider was reached through Bijection, and no case-scoped check receipt is attached.	Execution result unavailable

Hosted qualification: unestablished for this catalog configuration. Installed connections report their own identity, authority and current status.

Provider guidance

These recipes declare sources read by the native store transport: Azure Cosmos DB, Apache Cassandra, Elasticsearch, OpenSearch, Apache HBase (REST server), Neo4j (HTTP Query API), Redis keys, OPC UA servers and the AVEVA PI System through PI Web API. A document store record is { key, document }; a historian sample is { key, stream, time, value, status }; a discovery listing entry is { key, name, document }. The declaration names only what changes what a publication proves: which change history is followed, and deletions: { everyHours }, which looks every published key up by key at that interval and withdraws the keys the store answers absent for. A sweep is refused where the change history already carries deletions (Cosmos DB all versions and deletes, Neo4j change data capture), for OPC UA, and for PI points and values; a PI AF pass takes one. Endpoints, databases, containers, keyspaces, tables, indexes, labels, key properties, key patterns, nodes and points are binding setup; credentials are private connection configuration.

Recipe	Initial read	Changes	Deletions
`cosmosdb.container.changeFeed`	latest-version feed from the beginning	latest per key	sweep only
`cosmosdb.container.allVersions`	query snapshot	all versions and deletes feed	in the feed, within continuous-backup retention
`cassandra.table.scan`	paged scan at LOCAL_QUORUM	the next pass	sweep only
`elasticsearch.index.scan`, `opensearch.index.scan`	point-in-time scan	the next pass	sweep only
`hbase.table.scan` / `timeRange`	row scan	next pass / cell timestamp ranges	sweep only
`neo4j.nodes.cdc` / `watermark`	key-ordered snapshot	change data capture / watermark	CDC / sweep only
`redisKeys.keys.scan`	SCAN pass	next pass; notifications as hints	sweep, or a hinted key read absent
`opcua.nodes.browse`, `pi.points.list`	complete bounded listing	the next listing	omission from a complete listing
`pi.assets.list`	paged AF pass: database, elements, attributes	the next pass	sweep only (404 by WebId)
`opcua.values.history`, `pi.values.recorded`	archive from the installed start	values after the last one	none
`pi.values.channel`	archive from the installed start	archive backfill, then channel values within each window	none
`opcua.values.live`	none	subscription changes within each window	none

PI values read installed points (names on dataServer) and attributes (full AF attribute paths). A pi.values.channel connection is never continuous with the previous one: each first reads recorded values from the retained position, then follows the channel until its window ends or the channel closes, fails or stays silent past its heartbeat bound. PI recorded and channel sources may declare writes: { commands: { <name>: "create" } } (no marker): a create adds { value } at its key’s time (<WebId>|<time>|0) with updateOption=NoReplace and bufferOption=DoNotBuffer, and a lost answer is settled by reading the value back. OPC UA uses security policy None, so it reaches only loopback endpoints or a co-located gateway; signed and encrypted policies are not built. These registrations establish no provider check or hosted qualification; see the platform’s provider pages for what each reader proves and its limits.

Provider references

Provider references and prerequisites are maintained in the connector guidance above. No separate provider-documentation review date is recorded.

Source and evidence

Canonical connector source, native analysis and check receipts. Connector artifact: 905888ea276a2a28e36981a79e8142b2547ea451c3e63162cb6915ad4a7d1305. The resource declaration examples above are compiled and compared with exact catalog declarations. Recorded checks apply only to their source, declaration and runtime identities. Configure a connection and inspect its status to establish deployment-specific readiness.

Get Started

Use-case atlas

Platform

Data and Operations

Guides

Client Libraries

Tools

API Reference

Availability

Resources and modes

Index

scan: scan

Configuration: standard

Provider guidance

Provider references

Source and evidence

​Availability

​Resources and modes

​Index

​scan: scan

​Configuration: standard

​Provider guidance

​Provider references

​Source and evidence

Availability

Resources and modes

Index

scan: scan

Configuration: standard

Provider guidance

Provider references

Source and evidence