Add support for Rserve

.vscode/settings.json

@@ -18,5 +18,6 @@
     "python.testing.unittestEnabled": false,
     "python.testing.nosetestsEnabled": false,
     "python.testing.pytestEnabled": true,
-    "python.linting.pycodestyleEnabled": true
+    "python.linting.pycodestyleEnabled": true,
+    "python.pythonPath": "c:\\Users\\ogolovatyi\\.conda\\envs\\Python 3.7 (tornado 6)\\python.exe"
 }

CHANGELOG

@@ -1,5 +1,11 @@
 # Changelog
 
+## v1.2.0
+
+### Improvements
+
+- Added support for Rserve protocol.
+
 ## v1.1.0
 
 ### Improvements

docs/server-config.md

@@ -6,6 +6,9 @@
 
 - [Custom Settings](#custom-settings)
   * [Configuration File Content](#configuration-file-content)
+  * [Evaluation Engines](#evaluation-engines)
+    + [Python Evaluation Engine](#python-evaluation-engine)
+    + [Rserve Evaluation Engine](#rserve-evaluation-engine)
   * [Configuration File Example](#configuration-file-example)
 - [Configuring HTTP vs HTTPS](#configuring-http-vs-https)
 - [Authentication](#authentication)
@@ -14,6 +17,7 @@
   * [Adding an Account](#adding-an-account)
   * [Updating an Account](#updating-an-account)
   * [Deleting an Account](#deleting-an-account)
+  * [Endpoint Security](#endpoint-security)
 - [Logging](#logging)
   * [Request Context Logging](#request-context-logging)
 
@@ -91,77 +95,46 @@ at [`logging.config` documentation page](https://docs.python.org/3.6/library/log
   value - `30`. This timeout does not apply when evaluating models either
   through the `/query` method, or using the `tabpy.query(...)` syntax with
   the `/evaluate` method.
+- `TABPY_EVALUATE_WITH` - evaluation engine for user scripts. Supported
+  values are `Python` and `Rserve`. For specific configuration and other
+  details on each evaluation engine see [Evaluation Engines](#evaluation-engines)
+  section. Default value - `Python`.
+
+### Evaluation Engines
+
+TabPy supports evaluation user scripts with different evaluation engines.
+The engine can be specified with `TABPY_EVALUATE_WITH` configuration parameter.
+By default `Python` evaluation engine is used.
+
+#### Python Evaluation Engine
+
+User scripts are executed in the same Python instance where TabPy is running.
+No additional configuration parameters are needed for the engine (in
+`[TabPy]` section):
+
+- `EXTSVC_HOST` - Rserve host.
+- `EXTSVC_PORT` - Rserve port.
+
+Current limitations:
+
+- Authentication is not supported.
+
+#### Rserve Evaluation Engine
+
+TabPy forwards data and user script to specified
+[Rserve](https://www.rforge.net/Rserve/) instance.
+
+There are some additional configuration parameters needed for the engine:
+
+- Authentication is not supported.
 
 ### Configuration File Example
 
 **Note:** _Always use absolute paths for the configuration paths
 settings._
 
-```ini
-[TabPy]
-# TABPY_QUERY_OBJECT_PATH = /tmp/query_objects
-# TABPY_PORT = 9004
-# TABPY_STATE_PATH = <package-path>/tabpy/tabpy_server
-
-# Where static pages live
-# TABPY_STATIC_PATH = <package-path>/tabpy/tabpy_server/static
-
-# For how to configure TabPy authentication read
-# docs/server-config.md.
-# TABPY_PWD_FILE = /path/to/password/file.txt
-
-# To set up secure TabPy uncomment and modify the following lines.
-# Note only PEM-encoded x509 certificates are supported.
-# TABPY_TRANSFER_PROTOCOL = https
-# TABPY_CERTIFICATE_FILE = /path/to/certificate/file.crt
-# TABPY_KEY_FILE = /path/to/key/file.key
-
-# Log additional request details including caller IP, full URL, client
-# end user info if provided.
-# TABPY_LOG_DETAILS = true
-
-# Limit request size (in Mb) - any request which size exceeds
-# specified amount will be rejected by TabPy.
-# Default value is 100 Mb.
-# TABPY_MAX_REQUEST_SIZE_MB = 100
-
-# Configure how long a custom script provided to the /evaluate method
-# will run before throwing a TimeoutError.
-# The value should be a float representing the timeout time in seconds.
-# TABPY_EVALUATE_TIMEOUT = 30
-
-[loggers]
-keys=root
-
-[handlers]
-keys=rootHandler,rotatingFileHandler
-
-[formatters]
-keys=rootFormatter
-
-[logger_root]
-level=DEBUG
-handlers=rootHandler,rotatingFileHandler
-qualname=root
-propagete=0
-
-[handler_rootHandler]
-class=StreamHandler
-level=DEBUG
-formatter=rootFormatter
-args=(sys.stdout,)
-
-[handler_rotatingFileHandler]
-class=handlers.RotatingFileHandler
-level=DEBUG
-formatter=rootFormatter
-args=('tabpy_log.log', 'a', 1000000, 5)
-
-[formatter_rootFormatter]
-format=%(asctime)s [%(levelname)s] (%(filename)s:%(module)s:%(lineno)d): %(message)s
-datefmt=%Y-%m-%d,%H:%M:%S
-
-```
+Example file can be found at
+[https://github.com/tableau/TabPy/blob/master/tabpy/tabpy_server/common/default.conf](https://github.com/tableau/TabPy/blob/master/tabpy/tabpy_server/common/default.conf).
 
 ## Configuring HTTP vs HTTPS
 

docs/tabpy-virtualenv.md

@@ -51,7 +51,7 @@ To run TabPy in an Anaconda virtual environment follow the steps:
 in a Windows environment*
 
 1. For Windows open `Anaconda Prompt` from the Windows Start menu, for
-Linux and Mac run shell.
+   Linux and Mac run shell.
 
 2. Navigate to your home directory:
    1. On Windows run

setup.py

@@ -77,6 +77,7 @@ def read(fname):
             "numpy",
             "pandas",
             "pyopenssl",
+            "pyrserve",
             "python-dateutil",
             "requests",
             "scipy",

tabpy/VERSION

@@ -1 +1 @@
-1.1.0
+1.2.0

tabpy/models/deploy_models.py

@@ -1,6 +1,5 @@
 import os
 from pathlib import Path
-import pip
 import platform
 import subprocess
 import sys

tabpy/tabpy_server/app/app.py

@@ -1,16 +1,13 @@
-import concurrent.futures
 import configparser
 import logging
 from logging import config
-import multiprocessing
 import os
 import shutil
 import signal
 import sys
 import tabpy.tabpy_server
 from tabpy.tabpy import __version__
-from tabpy.tabpy_server.app.ConfigParameters import ConfigParameters
-from tabpy.tabpy_server.app.SettingsParameters import SettingsParameters
+from tabpy.tabpy_server.app.app_parameters import ConfigParameters, SettingsParameters
 from tabpy.tabpy_server.app.util import parse_pwd_file
 from tabpy.tabpy_server.management.state import TabPyState
 from tabpy.tabpy_server.management.util import _get_state_from_file
@@ -128,17 +125,10 @@ def try_exit(self):
         )
         logger.info("Done initializing TabPy.")
 
-        executor = concurrent.futures.ThreadPoolExecutor(
-            max_workers=multiprocessing.cpu_count()
-        )
-
         # initialize Tornado application
         _init_asyncio_patch()
         application = TabPyTornadoApp(
             [
-                # skip MainHandler to use StaticFileHandler .* page requests and
-                # default to index.html
-                # (r"/", MainHandler),
                 (
                     self.subdirectory + r"/query/([^/]+)",
                     QueryPlaneHandler,
@@ -152,11 +142,7 @@ def try_exit(self):
                     EndpointHandler,
                     dict(app=self),
                 ),
-                (
-                    self.subdirectory + r"/evaluate",
-                    EvaluationPlaneHandler,
-                    dict(executor=executor, app=self),
-                ),
+                (self.subdirectory + r"/evaluate", EvaluationPlaneHandler, dict(app=self)),
                 (
                     self.subdirectory + r"/configurations/endpoint_upload_destination",
                     UploadDestinationHandler,
@@ -213,6 +199,31 @@ def _set_parameter(self, parser, settings_key, config_key, default_val, parse_fu
         if not key_is_set:
             logger.debug(f"Parameter {settings_key} is not set")
 
+    def _parse_eval_with_config(self, parser):
+        """
+        Parse parameters specific for evaluation engine.
+        """
+
+        settings_parameters = []
+        evaluate_with = self.settings[SettingsParameters.EvaluateWith].lower()
+
+        if evaluate_with == "python":
+            # do nothing, Python evaluation requieres no additional parameters
+            pass
+        elif evaluate_with == "rserve":
+            settings_parameters = [
+                (SettingsParameters.ExtSvcHost, ConfigParameters.EXTSVC_HOST, None, None),
+                (SettingsParameters.ExtSvcPort, ConfigParameters.EXTSVC_PORT, None,
+                 parser.getint),
+            ]
+        else:
+            msg = f"Unknown evaluation engine '{evaluate_with}'"
+            logger.critical(msg)
+            raise RuntimeError(msg)
+
+        for setting, parameter, default_val, parse_function in settings_parameters:
+            self._set_parameter(parser, setting, parameter, default_val, parse_function)
+
     def _parse_config(self, config_file):
         """Provide consistent mechanism for pulling in configuration.
 
@@ -274,11 +285,16 @@ def _parse_config(self, config_file):
              "false", None),
             (SettingsParameters.MaxRequestSizeInMb, ConfigParameters.TABPY_MAX_REQUEST_SIZE_MB,
              100, None),
+            (SettingsParameters.EvaluateWith, ConfigParameters.TABPY_EVALUATE_WITH,
+             "Python", None),
         ]
 
         for setting, parameter, default_val, parse_function in settings_parameters:
             self._set_parameter(parser, setting, parameter, default_val, parse_function)
 
+        # parse evaluation engine specific parameters
+        self._parse_eval_with_config(parser)
+
         if not os.path.exists(self.settings[SettingsParameters.UploadDir]):
             os.makedirs(self.settings[SettingsParameters.UploadDir])
 
@@ -340,11 +356,6 @@ def _parse_config(self, config_file):
         logger.info(f"Call context logging is {call_context_state}")
 
     def _validate_transfer_protocol_settings(self):
-        if SettingsParameters.TransferProtocol not in self.settings:
-            msg = "Missing transfer protocol information."
-            logger.critical(msg)
-            raise RuntimeError(msg)
-
         protocol = self.settings[SettingsParameters.TransferProtocol]
 
         if protocol == "http":

tabpy/tabpy_server/app/app_parameters.py

@@ -0,0 +1,42 @@
+class ConfigParameters:
+    """
+    Configuration settings names
+    """
+
+    TABPY_PWD_FILE = "TABPY_PWD_FILE"
+    TABPY_PORT = "TABPY_PORT"
+    TABPY_QUERY_OBJECT_PATH = "TABPY_QUERY_OBJECT_PATH"
+    TABPY_STATE_PATH = "TABPY_STATE_PATH"
+    TABPY_TRANSFER_PROTOCOL = "TABPY_TRANSFER_PROTOCOL"
+    TABPY_CERTIFICATE_FILE = "TABPY_CERTIFICATE_FILE"
+    TABPY_KEY_FILE = "TABPY_KEY_FILE"
+    TABPY_PWD_FILE = "TABPY_PWD_FILE"
+    TABPY_LOG_DETAILS = "TABPY_LOG_DETAILS"
+    TABPY_STATIC_PATH = "TABPY_STATIC_PATH"
+    TABPY_MAX_REQUEST_SIZE_MB = "TABPY_MAX_REQUEST_SIZE_MB"
+    TABPY_EVALUATE_TIMEOUT = "TABPY_EVALUATE_TIMEOUT"
+    TABPY_EVALUATE_WITH = "TABPY_EVALUATE_WITH"
+    EXTSVC_HOST = "EXTSVC_HOST"
+    EXTSVC_PORT = "EXTSVC_PORT"
+
+
+class SettingsParameters:
+    """
+    Application (TabPyApp) settings names
+    """
+
+    TransferProtocol = "transfer_protocol"
+    Port = "port"
+    ServerVersion = "server_version"
+    UploadDir = "upload_dir"
+    CertificateFile = "certificate_file"
+    KeyFile = "key_file"
+    StateFilePath = "state_file_path"
+    ApiVersions = "versions"
+    LogRequestContext = "log_request_context"
+    StaticPath = "static_path"
+    MaxRequestSizeInMb = "max_request_size_in_mb"
+    EvaluateTimeout = "evaluate_timeout"
+    EvaluateWith = "evaluate_with"
+    ExtSvcHost = "extsvc_host"
+    ExtSvcPort = "extsvc_port"

tabpy/tabpy_server/common/default.conf

@@ -30,6 +30,22 @@
 # The value should be a float representing the timeout time in seconds.
 # TABPY_EVALUATE_TIMEOUT = 30
 
+# Configure evaluation engine for user scripts.
+# Supported values: Python, Rserve.
+# Default: Python.
+# TABPY_EVALUATE_WITH = Python
+
+# ======================================
+# Evaluation engine specific parameters.
+
+# External service host.
+# Used by: Rserve
+# EXTSVC_HOST = <host>
+
+# External service port.
+# Used by: Rserve
+# EXTSVC_PORT = <port>
+
 [loggers]
 keys=root
 

tabpy/tabpy_server/handlers/__init__.py

@@ -1,7 +1,5 @@
 from tabpy.tabpy_server.handlers.base_handler import BaseHandler
-from tabpy.tabpy_server.handlers.main_handler import MainHandler
 from tabpy.tabpy_server.handlers.management_handler import ManagementHandler
-
 from tabpy.tabpy_server.handlers.endpoint_handler import EndpointHandler
 from tabpy.tabpy_server.handlers.endpoints_handler import EndpointsHandler
 from tabpy.tabpy_server.handlers.evaluation_plane_handler import EvaluationPlaneHandler