Component
core/gateway/apisix
Apache APISIX, an API gateway. APISIX terminates client HTTP traffic on its proxy port and publishes that port's host, port and protocol. While nothing is wired to its models socket it routes by configuration held in the etcd store wired to its `etcd` socket, written through an Admin API reachable in-cluster only and authenticated by the key attached to `admin_key`; both are then required. It holds no state of its own. It also serves models: each one wired to its models socket gets a chat route, and a completion route unless a period budget or a prompt guard is attached, through APISIX's AI proxy, at /<alias>/v1 under the connection's alias, sent to the endpoint and model name the model publishes and presenting that connection's credential. Those routes are rendered into a file the gateway reads, so with a model wired it reads no etcd and runs no Admin API, and an `etcd` connection or an `admin_key` attachment halts the build. It publishes values.model's route as its endpoint and that alias as its model, so an agent's model socket takes it in place of a model; with nothing wired both are empty. On Kubernetes the Manifest chooses between two realizations. `apisix` synthesizes a workload with a hand-written configuration file and fetches nothing; it serves no model and halts on a models connection. `apisix-chart` renders the project's own Helm chart, which assembles that configuration from named settings; it serves the models, their callers and the controls attached to them. With `values.metrics_enabled` true, both serve the gateway's own metrics, asking no credential, and publish their address as `metrics_endpoint`; false, that key is empty.
来源
- 内容摘要
- sha256:4e088801181b19ecf4aab1e053711760de321bffb6b36bbf1961b9e1a0cc5274
- 发布时间
- 2026-11-10T10:11:40Z
- 签名身份
- https://token.actions.githubusercontent.com/https://github.com/aggrega-ai/aggrega-hub/.github/workflows/catalog-publish-sign.yml@refs/heads/main
签名
- 官方application/vnd.dev.sigstore.bundle.v0.3+json
https://token.actions.githubusercontent.com/https://github.com/aggrega-ai/aggrega-hub/.github/workflows/catalog-publish-sign.yml@refs/heads/main
sha256:9fb756fd0aeb866439da0421ca2c6162d4d0efee2944cf34a1abca4cfdb605a3
此版本的引用
oci://hub.aggrega.ai/component/core/gateway/apisix:1.0.0开发者参数
开发者在 Aggregate 中设置的值。
| 名称 | 类型 | 默认值 | 必需 | 描述 |
|---|---|---|---|---|
| model | string | — | 否 | The alias of a connection on the models socket that this gateway publishes as its model, and whose route its published endpoint addresses. It must be an alias a models connection carries, and the build halts naming this value otherwise; it is required while the models socket holds a connection, and omitted while it holds none. |
| metrics_enabled | boolean | false | 否 | Turns on the gateway's own metrics, in the Prometheus format, and publishes their address as `metrics_endpoint`; while false that key is empty. The port they are served on checks no caller, so anything that reaches it can read them. |
运维规格
运维人员按环境调整的基础设施规格。
| 名称 | 类型 | 默认值 | 必需 | 描述 |
|---|---|---|---|---|
| http_port | integer (min 1024, max 65535) | 9080 | 否 | Proxy data-plane port consumers connect to. |
| admin_port | integer (min 1024, max 65535) | 9180 | 否 | Admin API port, reachable in-cluster only. Read only while nothing is wired to the models socket; with a model wired the gateway runs no Admin API. |
| callers | array (min 1) | — | 否 | The callers the gateway serves models to, by name, each presenting its own key in the Authorization header as `Bearer <key>`, which is how an OpenAI client presents one. Every model route then refuses a request presenting no declared key with HTTP 401, removes the key from the request it forwards, and sends the caller's name to the model in an X-Consumer-Username header. A name takes letters, digits, hyphens and underscores, and is also the data key holding that caller's key in the Secret the caller_keys slot names. Read by `apisix-chart` while a model is wired; with nothing wired, or with `apisix`, the build halts on it. Omit it for open routes. |
| engine_version | string | 3.17.0-debian | 否 | Apache APISIX container image tag. Selects the image tag on container targets; the chart-backed realization pins its own image tag. |
| source_version | string | — | 否 | Version of the Helm chart the chart-backed realization pulls. Read only by that realization; the workload-synthesis realization pulls no source artifact and ignores it. Changing it away from the Recipe's default does not guarantee value-mapping compatibility with the new chart. |
| compute | object | — | 否 | — |
| overrides | object (open) | — | 否 | Provider-specific overrides. Accepts any nested structure. Each Recipe documents which keys it reads. |
输出契约
此组件通过 uses 拓扑向使用方发布的值。
| 名称 | 类型 | 默认值 | 必需 | 描述 |
|---|---|---|---|---|
| host | string | — | 是 | In-cluster DNS name of the Service fronting the proxy data plane. A consumer sends its client traffic here rather than to the service behind the gateway. |
| port | integer | — | 是 | TCP port that Service accepts proxied traffic on. It carries the data-plane port an operator set, never the Admin API port, which stays in-cluster. |
| protocol | string | — | 否 | How the published address is spoken to — `http`, so a consumer composes an address on that scheme. |
| endpoint | string | — | 否 | Full base URL of the OpenAI-compatible API the gateway serves for the model values.model names, with the host, the port and that alias's route already composed. A caller hands it to a client library as its base URL. Every other wired model is served at the same host and port under its own alias in the path. Empty while nothing is wired to the models socket. |
| model | string | — | 否 | The alias a consumer wired to the gateway sends as its model name: values.model. The route behind the endpoint asks for the model it was written for whatever name a request carries. Empty while nothing is wired to the models socket. |
| metrics_endpoint | string | — | 否 | The full address Prometheus scrapes this component's metrics at — scheme, host, port and path, `http://<host>:9091/apisix/prometheus/metrics`. Empty while values.metrics_enabled is false. |
uses 套接字
此组件用于建立连接的依赖套接字。通过某个套接字接入的每个连接,其目标都必须满足该套接字的标签选择器,且每个连接都可以在该套接字声明的槽位中携带附加项。必需槽位须由 Aggregate 或 Release 通过该套接字接入的每个连接来填充。
etcd
- 选择器
- component.aggrega.ai/type=stateful-server
- 注入
- env
此套接字未声明附加槽位,因此经由它的连接不携带任何附加项。
models
- 选择器
- component.aggrega.ai/category=ai-model
- 注入
- —
经由此套接字的每个连接可填充的槽位
| 槽位 | 选择器 | 必需 | 注入 | 敏感 | 描述 |
|---|---|---|---|---|---|
| credential | feature.aggrega.ai/kind=secret | 否 | env | 是 | Optional credential the gateway presents to the model this connection reaches, as a bearer token. When attached, the Recipe delivers it as AI_PROXY_<ALIAS>_API_KEY, the connection's alias in capitals with every character other than a letter or a digit as an underscore, by a Kubernetes secretKeyRef, composes the bearer token AI_PROXY_<ALIAS>_AUTHORIZATION from it, and the connection's routes read that token through APISIX's $env:// reference, naming the variable rather than the value. Unattached, the route adds no credential. |
cache_store
- 选择器
- datastore.aggrega.ai/engine=kv
- 注入
- —
经由此套接字的每个连接可填充的槽位
| 槽位 | 选择器 | 必需 | 注入 | 敏感 | 描述 |
|---|---|---|---|---|---|
| credential | feature.aggrega.ai/kind=secret | 否 | env | 是 | Optional password the gateway's semantic cache presents to the store this connection reaches. When attached while core/semantic-caching is attached, the Recipe delivers it as CACHE_STORE_PASSWORD by a Kubernetes secretKeyRef, and the cache reads it through APISIX's $env:// reference, naming the variable rather than the value; attach a credential holding the password the store itself requires, which must publish the name of a Secret and the data key holding it there. Unattached, the cache connects without a password, which suits a store with none. |
附加槽位
Feature 附加到此组件时所用的命名绑定点。槽位的选择器限定哪个 Feature 可以填充该槽位,其 inject 列表则指明 Feature 的内容可以如何交付。必需槽位必须在每个实例上完成绑定,可以通过实例上的附加,也可以通过其上层作用域的附加;敏感槽位则声明为用于处理机密内容。Feature 也可以以任何槽位都未声明的名称进行附加。
| 槽位 | 选择器 | 必需 | 注入 | 敏感 | 描述 |
|---|---|---|---|---|---|
| admin_key | feature.aggrega.ai/kind=secret | 否 | env | 是 | The key authenticating the gateway's Admin API, which a caller presents in the X-API-KEY header. The attached credential must publish the name of a Secret and the data key holding the key there; each Recipe delivers the key from that Secret to the gateway as an environment variable, and a credential publishing no such locator halts the build. Required while nothing is wired to the models socket, with no default: with no credential attached then, the build halts. While a model is wired the gateway runs no Admin API, and an attachment here halts the build. The key must not be empty, which stops the gateway at start-up, and must not be a value APISIX reads as a number or as true or false, such as an all-digit key, because APISIX then never matches it against the key a request presents. |
| caller_keys | feature.aggrega.ai/kind=secret | 否 | env | 是 | The Secret holding the key each caller infra.callers declares presents to the gateway, one data key per caller, the data key being the caller's name. The attached credential must publish the Secret's name, as core/referenced-secret-set does, and a credential publishing none halts the build. The chart Recipe delivers each key as CALLER_<NAME>_KEY, the caller's name in capitals with every character other than a letter or a digit as an underscore, by a Kubernetes secretKeyRef, and composes CALLER_<NAME>_AUTHORIZATION, the whole Authorization header a caller presents, from it; the caller's consumer in the routes file reads that header through APISIX's $env:// reference. Required while infra.callers declares a caller, and refused while it declares none. A caller with no data key of its name in the Secret leaves the gateway unable to start, since Kubernetes starts no container whose secretKeyRef names a missing key; that is caught at start, not by the build. |
自注入允许列表
env
关系
与其他已发布实体之间的类型化边。没有已发布目标的引用会呈现为断开的边,而不是链接。
使用
- 使用使用套接字 «etcd»
component.aggrega.ai/type=stateful-server
未声明附加槽位。
core/ai-agent/dify-plugin-daemon v1.0.0 →core/ai-tool/filesystem v1.0.0 →core/datastore/chroma v1.0.0 →core/datastore/cognee v1.0.0 →core/datastore/documentdb v1.0.0 →core/datastore/elasticsearch v1.0.0 →core/datastore/etcd v1.0.0 →core/datastore/ferretdb v1.0.0 →core/datastore/mariadb v1.0.0 →core/datastore/milvus v1.0.0 →core/datastore/mysql v1.0.0 →core/datastore/mysql-innodb-cluster v1.0.0 →core/datastore/opensearch v1.0.0 →core/datastore/pgvector v1.0.0 →core/datastore/postgres v1.0.0 →core/datastore/qdrant v1.0.0 →core/datastore/timescaledb v1.0.0 →core/datastore/weaviate v1.0.0 →core/devtool/harbor v1.0.0 →core/messaging/kafka v1.0.0 →core/messaging/rabbitmq v1.0.0 →core/observability/fluentd v1.0.0 →core/observability/jaeger v1.0.0 →core/observability/loki v1.0.0 →core/observability/prometheus v1.0.0 →core/observability/tempo v1.0.0 →core/security/openbao v1.0.0 → - 使用使用套接字 «models»
component.aggrega.ai/category=ai-model
附加槽位:credential
core/ai-model/ai21 v1.0.0 →core/ai-model/anthropic v1.0.0 →core/ai-model/azure-managed v1.0.0 →core/ai-model/azure-openai v1.0.0 →core/ai-model/azure-openai-embedding v1.0.0 →core/ai-model/bedrock v1.0.0 →core/ai-model/bedrock-embedding v1.0.0 →core/ai-model/cohere v1.0.0 →core/ai-model/databricks v1.0.0 →core/ai-model/docker-model-runner v1.0.0 →core/ai-model/gemini v1.0.0 →core/ai-model/huggingface v1.0.0 →core/ai-model/litellm v1.0.0 →core/ai-model/mistral v1.0.0 →core/ai-model/nemoguard-content-safety v1.0.0 →core/ai-model/nvidia-api v1.0.0 →core/ai-model/nvidia-nim v1.0.0 →core/ai-model/nvidia-nim-operator v1.0.0 →core/ai-model/ollama v1.0.0 →core/ai-model/openai v1.0.0 →core/ai-model/openai-compatible v1.0.0 →core/ai-model/openai-embedding v1.0.0 →core/ai-model/replicate v1.0.0 →core/ai-model/sagemaker v1.0.0 →core/ai-model/tei v1.0.0 →core/ai-model/tgi v1.0.0 →core/ai-model/vertex-ai v1.0.0 →core/ai-model/vertex-ai-embedding v1.0.0 →core/ai-model/vllm v1.0.0 → - 使用使用套接字 «cache_store»
datastore.aggrega.ai/engine=kv
附加槽位:credential
被使用
使用套接字 «model»
使用套接字 «model»
使用套接字 «model»
使用套接字 «model»
使用套接字 «model»
使用套接字 «model»
使用套接字 «model»
使用套接字 «model»
使用套接字 «model»
使用套接字 «model»
使用套接字 «model»
使用套接字 «model»
使用套接字 «model»
使用套接字 «model»
使用套接字 «model»
使用套接字 «model»
使用套接字 «model»
使用套接字 «model»
使用套接字 «model»
使用套接字 «model»
使用套接字 «model»
使用套接字 «model»
使用套接字 «model»
使用套接字 «model»
使用套接字 «model»
使用套接字 «model»
使用套接字 «model»
使用套接字 «model»
使用套接字 «model»
使用套接字 «model»
使用套接字 «model»
使用套接字 «model»
使用套接字 «model»
使用套接字 «target_model»
使用套接字 «extraction_model»
使用套接字 «embedding_model»
使用套接字 «upstream»
使用套接字 «forward_target»
使用套接字 «scrape_target»
使用套接字 «store»
适用于
适用于
适用于
适用于
适用于
适用于
适用于
适用于
适用于
适用于
适用于
适用于
适用于
适用于
适用于
适用于
实现
实现