[GSK-1698] Remove legacy ML worker info and properties

backend/src/main/java/ai/giskard/config/ApplicationProperties.java

@@ -16,8 +16,6 @@
 @Getter
 @ConfigurationProperties(prefix = "giskard", ignoreUnknownFields = false)
 public class ApplicationProperties {
-    private String mlWorkerHost;
-    private int mlWorkerPort;
     private int externalMlWorkerEntrypointPort;
     private String externalMlWorkerEntrypointHost;
     private boolean externalMlWorkerEnabled;

backend/src/main/resources/config/application.yml

@@ -162,10 +162,6 @@ info:
   display-ribbon-on-profiles: 'dev'
 
 giskard:
-  ml-worker-host: localhost
-  ml-worker-port: 50051
-  external-ml-worker-entrypoint-port: 40051
-  external-ml-worker-enabled: true
   api-token-validity-in-days: 90
   invitation-token-validity-in-days: 3
   borderLineThreshold: 0.1

backend/src/test/resources/config/application.yml

@@ -80,8 +80,6 @@ server:
 giskard:
   api-token-validity-in-days: 90
   invitation-token-validity-in-days: 3
-  ml-worker-host: localhost
-  ml-worker-port: 50051
   home: ${user.home}/giskard-home
   auth: true
   email-from: giskard@localhost

python-client/docs/guides/installation_app/index.rst

@@ -62,7 +62,7 @@ You can either install and run the server **locally** or on an **external server
 
       Installing Giskard in the cloud is preferable if you want to use the **collaborative** features of Giskard: collect feedback on your model from your team, share your Quality Assurance results, save and provide all your custom tests to your team, etc.
 
-      Since Giskard uses 2 TCP ports: ``19000`` and ``40051``, **make sure that these two ports are open** on the cloud instances where Giskard is installed. For step-by-step installation steps in the cloud, please go to the `AWS <install_aws/index/index.md>`_, `GCP <install_gcp/index.md>`_, and `Azure <install_azure/index.md>`_ installation pages.
+      Since Giskard uses a TCP ports: ``19000``, **make sure that the port is open** on the cloud instances where Giskard is installed. For step-by-step installation steps in the cloud, please go to the `AWS <install_aws/index/index.md>`_, `GCP <install_gcp/index.md>`_, and `Azure <install_azure/index.md>`_ installation pages.
 
       .. toctree::
          :maxdepth: 1
@@ -123,8 +123,6 @@ Giskard executes your model using an worker that runs directly the model in your
 
          .. code-block:: sh
 
-            %env GSK_EXTERNAL_ML_WORKER_HOST=4.tcp.ngrok.io
-            %env GSK_EXTERNAL_ML_WORKER_PORT=10853
             %env GSK_API_KEY=YOUR_API_KEY
             !giskard worker start -d -k YOUR_TOKEN -u https://e840-93-23-184-184.ngrok-free.app
 

python-client/docs/guides/installation_app/install_aws/index.md

@@ -10,7 +10,7 @@
 * **Application and OS image**: Select the default Ubuntu server 22.04 LTS 64-bit (x86)
 * **Instance type**: We recommend you to choose at least a `t2.large` instance type (2vCPU, 8GB memory)
 * **Key pair**: Choose your usual key pair. If you don't have one, go to the [Amazon document](https://docs.aws.amazon.com/AWSEC2/latest/UserGuide/create-key-pairs.html) to create the right one
-* **Network settings**: You need to **open the ports `19000` and `40051` (TCP connection)** to access the Giskard frontend (port `19000`) and upload your model (port `40051`). To do so, click on `Edit` and add the following security groups:
+* **Network settings**: You need to **open the port `19000` to access the Giskard frontend and upload your model. To do so, click on `Edit` and add the following security groups:
 
 <figure><img src="../../.gitbook/assets/image (1) (1) (1).png" alt=""><figcaption></figcaption></figure>
 

python-client/docs/guides/installation_app/install_gcp/index.md

@@ -10,11 +10,10 @@ Installing Giskard in GCP enables you to inspect & test models that you created
     2. Choose `Allow full access to all Cloud APIs`
     3. In the firewall section, allow **HTTP** and **HTTPS** traffic
 3. Connect to your VM in SSH by opening a browser window
-4. Create a firewall rule to open ports `19000` and `40051` port of the Giskard instance. Here is the command line that you can execute in the terminal opened by your SSH connection:
+4. Create a firewall rule to open the `19000` port of the Giskard instance. Here is the command line that you can execute in the terminal opened by your SSH connection:
 
 ```bash
 gcloud compute firewall-rules create giskard-main --allow tcp:19000
-gcloud compute firewall-rules create giskard-worker --allow tcp:40051
 ```
 
 :::{warning}
@@ -32,7 +31,6 @@ Creating the firewall rules can also be done **through UI** in the `VPC Network`
 * In `source IPv4 ranges,` select `0.0.0.0/0`
 * In `Protocols and ports`, select `Specified protocols and ports`
 * Then select `TCP`, and type `19000`
-* `Repeat the same steps to open port 40051`
 :::
 
 ### 2. Install Giskard in the GCP VM

python-client/giskard/commands/cli_server.py

@@ -3,7 +3,6 @@
 import time
 from pathlib import Path
 from typing import Optional
-from urllib.parse import urlparse
 
 import click
 import docker
@@ -41,9 +40,7 @@ def create_docker_client() -> DockerClient:
         exit(1)
 
 
-@click.group(
-    "server", help="Giskard UI management", context_settings={"show_default": True}
-)
+@click.group("server", help="Giskard UI management", context_settings={"show_default": True})
 def server() -> None:
     """
     Giskard UI management
@@ -88,9 +85,7 @@ def get_container(version=None, quit_if_not_exists=True) -> Optional[Container]:
         return create_docker_client().containers.get(name)
     except NotFound:
         if quit_if_not_exists:
-            logger.error(
-                f"Container {name} could not be found. Run `giskard server start` to create the container"
-            )
+            logger.error(f"Container {name} could not be found. Run `giskard server start` to create the container")
             raise click.Abort()
         else:
             return None
@@ -128,7 +123,6 @@ def _start(attached=False, version=None):
 
     settings = _get_settings() or {}
     port = settings.get("port", 19000)
-    ml_worker_port = settings.get("ml_worker_port", 40051)
 
     version = get_version(version)
 
@@ -143,17 +137,15 @@ def _start(attached=False, version=None):
             get_image_name(version),
             detach=not attached,
             name=get_container_name(version),
-            ports={7860: port, 40051: ml_worker_port},
+            ports={7860: port},
             volumes={home_volume.name: {"bind": "/home/giskard/datadir", "mode": "rw"}},
         )
     container.start()
 
     up = _wait_backend_ready(port)
 
     if up:
-        logger.info(
-            f"Giskard Server {version} started. You can access it at http://localhost:{port}"
-        )
+        logger.info(f"Giskard Server {version} started. You can access it at http://localhost:{port}")
     else:
         logger.warning(
             "Giskard backend takes unusually long time to start, "
@@ -190,17 +182,13 @@ def _fetch_latest_tag() -> str:
     """
     Returns: the latest tag from the Docker Hub API. Format: vX.Y.Z
     """
-    response = requests.get(
-        "https://hub.docker.com/v2/namespaces/giskardai/repositories/giskard/tags?page_size=10"
-    )
+    response = requests.get("https://hub.docker.com/v2/namespaces/giskardai/repositories/giskard/tags?page_size=10")
     response.raise_for_status()
     json_response = response.json()
     latest_tag = "latest"
     latest = next(i for i in json_response["results"] if i["name"] == latest_tag)
     latest_version_image = next(
-        i
-        for i in json_response["results"]
-        if ((i["name"] != latest_tag) and (i["digest"] == latest["digest"]))
+        i for i in json_response["results"] if ((i["name"] != latest_tag) and (i["digest"] == latest["digest"]))
     )
 
     tag = latest_version_image["name"]
@@ -236,9 +224,7 @@ def _expose(token):
     container = get_container()
     if container:
         if container.status != "running":
-            print(
-                "Error: Giskard server is not running. Please start it using `giskard server start`"
-            )
+            print("Error: Giskard server is not running. Please start it using `giskard server start`")
             raise click.Abort()
     else:
         raise click.Abort()
@@ -249,29 +235,17 @@ def _expose(token):
     if token:
         ngrok.set_auth_token(token)
 
-    http_tunnel = ngrok.connect(
-        19000, "http", pyngrok_config=None if token else PyngrokConfig(region="us")
-    )
-    tcp_tunnel = ngrok.connect(
-        40051, "tcp", pyngrok_config=None if token else PyngrokConfig(region="eu")
-    )
-
-    # Only split the last ':' in case the URL contains a port
-    tcp_addr = urlparse(tcp_tunnel.public_url)
+    http_tunnel = ngrok.connect(19000, "http", pyngrok_config=None if token else PyngrokConfig(region="us"))
 
     print("Giskard Server is now exposed to the internet.")
-    print(
-        "You can now upload objects to the Giskard Server using the following client: \n"
-    )
+    print("You can now upload objects to the Giskard Server using the following client: \n")
 
     print(
         f"""token=...
 client = giskard.GiskardClient(\"{http_tunnel.public_url}\", token)
 
 # To run your model with the Giskard Server, execute these three lines on Google Colab:
 
-%env GSK_EXTERNAL_ML_WORKER_HOST={tcp_addr.hostname}
-%env GSK_EXTERNAL_ML_WORKER_PORT={tcp_addr.port}
 %env GSK_API_KEY=...
 !giskard worker start -d -u {http_tunnel.public_url}"""
     )
@@ -294,9 +268,7 @@ def _expose(token):
     default=False,
     help="Starts the server and attaches to it, displaying logs in console.",
 )
-@click.option(
-    "--version", "version", required=False, help="Version of Giskard server to start"
-)
+@click.option("--version", "version", required=False, help="Version of Giskard server to start")
 @common_options
 def start(attached, version):
     """\b
@@ -373,12 +345,8 @@ def restart(service, hard):
                 logger.info(f"Container {container.name} restarted")
         else:
             if service:
-                logger.info(
-                    f"Restarting service {service} in {container.name} container"
-                )
-                command = (
-                    f"supervisorctl -c /opt/giskard/supervisord.conf restart {service}"
-                )
+                logger.info(f"Restarting service {service} in {container.name} container")
+                command = f"supervisorctl -c /opt/giskard/supervisord.conf restart {service}"
             else:
                 logger.info(f"Restarting all services in {container.name} container")
                 command = "supervisorctl -c /opt/giskard/supervisord.conf restart all"
@@ -455,15 +423,11 @@ def diagnose(local_dir):
     analytics.track("giskard-server:diagnose")
     out_dir = Path(local_dir)
     assert out_dir.is_dir(), "'output' should be an existing directory"
-    bits, _ = get_container().get_archive(
-        "/home/giskard/datadir/run", encode_stream=True
-    )
+    bits, _ = get_container().get_archive("/home/giskard/datadir/run", encode_stream=True)
     from datetime import datetime
 
     now = datetime.now().strftime("%Y%m%d_%H%M%S_%f")
-    out_file = (
-        out_dir / f"giskard-diagnose-{get_version().replace('.', '_')}-{now}.tar.gz"
-    )
+    out_file = out_dir / f"giskard-diagnose-{get_version().replace('.', '_')}-{now}.tar.gz"
     with open(out_file, "wb") as f:
         for chunk in bits:
             f.write(chunk)
@@ -514,9 +478,7 @@ def status():
     analytics.track("giskard-server:status")
     app_settings = _get_settings()
     if not app_settings:
-        logger.info(
-            "Giskard Server is not installed. Install using `giskard server start`"
-        )
+        logger.info("Giskard Server is not installed. Install using `giskard server start`")
         return
     else:
         version = app_settings["version"]
@@ -532,21 +494,13 @@ def status():
     if container:
         if container.status == "running":
             logger.info(f"Container {container.name} status:")
-            print(
-                get_container()
-                .exec_run("supervisorctl -c /opt/giskard/supervisord.conf")
-                .output.decode()
-            )
+            print(get_container().exec_run("supervisorctl -c /opt/giskard/supervisord.conf").output.decode())
         else:
-            logger.info(
-                f"Container {container.name} isn't running ({container.status})"
-            )
+            logger.info(f"Container {container.name} isn't running ({container.status})")
 
 
 @server.command("clean")
-@click.option(
-    "--data", "delete_data", is_flag=True, help="Delete user data (giskard-home volume)"
-)
+@click.option("--data", "delete_data", is_flag=True, help="Delete user data (giskard-home volume)")
 @common_options
 def clean(delete_data):
     """\b

python-client/giskard/settings.py

@@ -18,13 +18,10 @@ def expand_env_var(env_var: Optional[str]) -> Optional[str]:
 
 class Settings(BaseSettings):
     home: str = "~/giskard-home"
-    port: int = 50051
     ws_port: int = 9000
     ws_path: str = "/ml-worker"
     host: str = "localhost"
     max_workers: int = 10
-    max_send_message_length_mb: int = 1024
-    max_receive_message_length_mb: int = 1024
     loglevel = "INFO"
     cache_dir: str = "cache"
     disable_analytics = False

scripts/ngrok.yml

@@ -10,7 +10,4 @@ authtoken: YOUR_AUTHTOKEN_HERE
 tunnels:
   backend:
     proto: http
-    addr: 19000
-  mlworker:
-    proto: tcp
-    addr: 40051
+    addr: 19000