- Deployment:
    - deployment/index.md
    - deployment/versions.md
    - deployment/https.md
    - deployment/manually.md
    - deployment/concepts.md
    - deployment/cloud.md
    - deployment/server-workers.md
    - deployment/docker.md
  - How To - Recipes:
    - how-to/index.md
    - how-to/general.md
    - how-to/graphql.md
    - how-to/custom-request-and-route.md
    - how-to/conditional-openapi.md

                    <mxGeometry x="420" y="280" width="920" height="670" as="geometry"/>
                </mxCell>
                <mxCell id="3" value="&lt;font face=&quot;Roboto&quot;&gt;&lt;span style=&quot;font-size: 24px&quot;&gt;Server&lt;/span&gt;&lt;/font&gt;" style="text;html=1;strokeColor=none;fillColor=none;align=center;verticalAlign=middle;whiteSpace=wrap;rounded=0;strokeWidth=3;fontFamily=Roboto Mono, mono;FType=g;" parent="1" vertex="1">

    data.write("8180".decodeHex())
    assertFailsWith<ProtocolException> {
      clientReader.processNextFrame()
    }.also { expected ->
      assertThat(expected.message)
        .isEqualTo("Server-sent frames must not be masked.")
    }
  }

  @Test fun controlFramePayloadMax() {
    data.write("8a7e007e".decodeHex())
    assertFailsWith<ProtocolException> {
      clientReader.processNextFrame()

    val address = factory.newAddress()
    proxySelector.proxies.add(Proxy.NO_PROXY)
    val routeSelector = newRouteSelector(address)
    proxySelector.assertRequests(address.url.toUri())

    // Only the origin server will be attempted.
    assertThat(routeSelector.hasNext()).isTrue()
    dns[uriHost] = dns.allocate(1)
    val selection = routeSelector.next()

    --hash=sha256:a6df30f272c08bf8be66e0775fad862005d950a6b8449b94f7c788731d70ecd7
    # via -r ci/official/requirements_updater/requirements.in
tensorboard-data-server==0.7.2 \
    --hash=sha256:7e0610d205889588983836ec05dc098e80f97b7e7bbff7e994ebb78f578d0ddb \
    --hash=sha256:9fe5d24221b29625dbc7328b0436ca7fc1c23de4acf4d272f1180856e32f9f60 \

    --hash=sha256:a6df30f272c08bf8be66e0775fad862005d950a6b8449b94f7c788731d70ecd7
    # via -r ci/official/requirements_updater/requirements.in
tensorboard-data-server==0.7.2 \
    --hash=sha256:7e0610d205889588983836ec05dc098e80f97b7e7bbff7e994ebb78f578d0ddb \
    --hash=sha256:9fe5d24221b29625dbc7328b0436ca7fc1c23de4acf4d272f1180856e32f9f60 \

    --hash=sha256:a6df30f272c08bf8be66e0775fad862005d950a6b8449b94f7c788731d70ecd7
    # via -r ci/official/requirements_updater/requirements.in
tensorboard-data-server==0.7.2 \
    --hash=sha256:7e0610d205889588983836ec05dc098e80f97b7e7bbff7e994ebb78f578d0ddb \
    --hash=sha256:9fe5d24221b29625dbc7328b0436ca7fc1c23de4acf4d272f1180856e32f9f60 \

---

"_We adopted the **FastAPI** library to spawn a **REST** server that can be queried to obtain **predictions**. [for Ludwig]_"

* PostgreSQL
* MySQL
* SQLite
* Oracle
* Microsoft SQL Server, etc.

In this example, we'll use **SQLite**, because it uses a single file and Python has integrated support. So, you can copy this example and run it as is.

Later, for your production application, you might want to use a database server like **PostgreSQL**.

/// tip

   * as when it was first created.
   *
   * <p>The returned {@code RateLimiter} is intended for cases where the resource that actually
   * fulfills the requests (e.g., a remote server) needs "warmup" time, rather than being
   * immediately accessed at the stable (maximum) rate.
   *
   * <p>The returned {@code RateLimiter} starts in a "cold" state (i.e. the warmup period will