</error-page>
  <error-page>
    <error-code>404</error-code>
    <location>/WEB-INF/view/error/redirect.jsp?type=notFound</location>
  </error-page>
  <error-page>
    <error-code>408</error-code>
    <location>/WEB-INF/view/error/redirect.jsp?type=logOut</location>
  </error-page>
  <error-page>
    <error-code>429</error-code>
    <location>/WEB-INF/view/error/redirect.jsp?type=busy</location>

    val response = client.newCall(request).execute()
    assertThat(response.code).isEqualTo(408)
    assertThat(response.body.string()).isEqualTo("You took too long!")
  }

  @Test
  fun maxClientRequestTimeoutRetries() {
    server.enqueue(
      MockResponse
        .Builder()
        .code(408)
        .addHeader("Connection", "Close")
        .body("You took too long!")

    if (responseCode == HttpURLConnection.HTTP_CLIENT_TIMEOUT) {
      // 408's are a bit of an outlier because we may repeat the request if we encounter this
      // response code. In this scenario, there are 2 responses: the initial 408 and then the 200
      // because of the retry. We just want to ensure the initial 408 isn't cached.
      expectedResponseCode = 200
      server.enqueue(
        MockResponse

            return LlmException.ERROR_AUTH;
        }
        if (statusCode == 404) {
            return LlmException.ERROR_MODEL_NOT_FOUND;
        }
        if (statusCode == 408) {
            return LlmException.ERROR_TIMEOUT;
        }
        if (statusCode == 502 || statusCode == 503) {
            return LlmException.ERROR_SERVICE_UNAVAILABLE;
        }