Development (#25)

arseniiarsenii · PaTara43 · web-flow · commit 40926564d8ac · 2022-09-02T15:23:56.000+03:00
* longer wait time * speed up unit retrieval with serverside stage aggregation * added execution timing for frequently used db methods * implemented exec timer for async functions * increased units pending revision fetching speed 3x * access collections with properties * removed unused mount * Reworked unit upload and update for x3 speed increase. Removed unused methods. * Version bump * Add WS endpoint for workbench status streaming * Add optional dotenv parsing * Implement SSE stream instead of WS * Fixed type annotation * update black version * reworked connection close pipeline * improved logging * Implement background printing tasks * Update docker-compose.yml * remove .gitignore * mark endpoint as deprecated * optimize txn hash pushing to db * removed polling and optimized state push using async event * bugfix: thread wouldn't unblock until txn posted. implemented async wrapper * implemented single field updates * async short url generator. up to 4s faster passport saving * updated packages * version bump * cached files auto remove * added some docker volumes * added db connection close on quit * update .gitignore * removed duplicate code * removed redundant db field * refactored unit retrieval * implemented workbench shutdown sequence * renamed actions * implemented multiplatform Dockerfile * multiplatform image building * bugfix * bugfix * removed arm v7 form the list of platforms * added comments * updated robonomics-interface version * updated Dockerfiles to matched unified code style * updated Dockerfile names * updated action to reuse in other repos * update type hints to 3.10 format * minor changes * bugfix * dependency update * update base version * update Python version in action * bugfix * updated dependencies * action bugfix attempt * added type annotation * update mypy config * update dep versions in requirements.txt * return caching in action * return caching in action * moved type alias * added back union checking plugin * filter out AssertionErrors * replace variable returns with HTTPExceptions * add black daemon * minor refactoring * fixed typos * fixed typo * minor update to passport code logic * filter access logs for certain endpoints * add healthcheck * added connectivity checks for external services * fix typo * bugfix * Update _Barcode.py (#22) * getting component staged for revision from the parent's id (MAP-257) * rewritten pre-commit config * isort * improved type annotations * bugfix * bugfix MAP-332 * Added ECS logging support for the ELK stack integration * slightly modified README * ECS logs now written to file instead of console * experimental: cache Docker layers * improved logging * updated pc hooks * Implemented SSE notifications * Optimized imports * Allow assigning built units with unreleased passport (MAP-355) * Tagged endpoints * update deps * update deps * notify user of a forbidden transition * add error notifications * fix mypy error * replaced concat with fstring * extracted singular connection func * add fstring instead of concat * added more checks and error notifications * added more checks and error notifications * added datalog success notification * version bump * bugfix * add more notifications * replaced json err responses with HTTP exceptions * more notifications and minor improvements * bring back bg datalog posting * reworked messenger interface, suppressed Runtime Error (temp fix) * Fixed 'RuntimeError: no running event loop' * User notifications translated to russian * Improved error notifications * Minor improvements to Messenger.py * Improved reliability for unit wrap up operation * Improved reliability for operation ending * Minor fixes * Updated dependencies * Added more user notifications * Improved image annotations for passport QR codes * Add unit's combined assembly time to the passport * Faster operation ending on composite units * Attempt to fix QR code side logos * Reworked unit component assignment logic * Avoid reassigning the same unit twice * Remove exception info from the printing error notifications * add prometheus metrics * renamed env var for mongodb uri * version bump * add docker secrets parsing * insert exception message into a metric label * reworked Robonomics async wrapper * version bump * add custom tracked exception * remove dependency compilation stage as target binaries are now available * bugfix * add version posting to prometheus * simplify error messages * Added 6 new metrics for basic worker productivity stats * bugfix * Fixed app version reporting inside Docker * Updated dependencies * Bugfix in version file parsing * Attempt connection error fix * Now sorting schemas in alphabetical order * Dependencies Updated * Removed unused Dockerfile Co-authored-by: Pavel Tarasov <34866682+PaTara43@users.noreply.github.com>
diff --git a/Dockerfile b/Dockerfile
@@ -1,39 +1,23 @@
-# There are no binaries of this packages as of today for ARM platforms,
-# so compilation is required
-FROM python:3.10 as dependency-compilation
-WORKDIR /tmp
-RUN apt-get update && apt-get install -y\
-    build-essential\
-    libssl-dev\
-    libffi-dev\
-    python3-dev
-RUN curl https://sh.rustup.rs -sSf | sh -s -- -y
-RUN /root/.cargo/bin/rustup default nightly
-RUN pip install --upgrade pip
-RUN . /root/.cargo/env && pip install py-bip39-bindings
-RUN . /root/.cargo/env && pip install py-ed25519-bindings
-RUN . /root/.cargo/env && pip install py-sr25519-bindings
-RUN . /root/.cargo/env && pip install cryptography
-
 # At this stage we convert Poetry's dependency file into a more traditional
 # requirements.txt to avoid installing Poetry into the final container.
 # Although very slow, this cannot be skipped, as we need to resolve dependencies
 # for the exact platform the client is using.
 FROM python:3.10 as requirements-stage
 WORKDIR /tmp
-COPY --from=dependency-compilation /root/.cache/pip /root/.cache/pip
+RUN pip install --upgrade pip
 RUN pip install poetry
 COPY ./pyproject.toml ./poetry.lock* /tmp/
 RUN poetry export -f requirements.txt --output requirements.txt --without-hashes
+RUN poetry version | grep -o [0-9.]* > version.txt
 
 # Final container build. Uses pre-compiled dependencies and requirements.txt
 # obtained in the previous steps
 FROM python:3.10
 WORKDIR /src
 COPY --from=requirements-stage /tmp/requirements.txt /src/requirements.txt
-COPY --from=dependency-compilation /root/.cache/pip /root/.cache/pip
 RUN pip install --no-cache-dir --upgrade -r /src/requirements.txt
 COPY ./src /src
+COPY --from=requirements-stage /tmp/version.txt /src/version.txt
 HEALTHCHECK --interval=5s --timeout=3s --start-period=5s --retries=12 \
     CMD curl --fail http://localhost:5000/docs || exit 1
 ENTRYPOINT ["uvicorn", "app:app", "--host", "0.0.0.0", "--port", "5000"]
diff --git a/Dockerfile-quick b/Dockerfile-quick
diff --git a/poetry.lock b/poetry.lock
diff --git a/pyproject.toml b/pyproject.toml
@@ -1,6 +1,6 @@
 [tool.poetry]
 name = "Feecc-Workbench-Daemon"
-version = "1.3.1"
+version = "1.3.2"
 description = "Workbench software for the Feecc QA system"
 authors = ["arseniiarsenii <arseniivelichko2@gmail.com>"]
 license = "Apache 2.0"
@@ -12,36 +12,36 @@ PyYAML = "^6.0"
 brother-ql = "^0.9.4"
 qrcode = "^7.3.1"
 python-barcode = "^0.14.0"
-fastapi = "^0.78.0"
-uvicorn = "^0.17.6"
+fastapi = "^0.81.0"
+uvicorn = "^0.18.3"
 dnspython = "^2.2.1"
 loguru = "^0.6.0"
 motor = "^3.0.0"
 httpx = "^0.23.0"
 environ-config = "^22.1.0"
-yarl = "^1.7.2"
-sse-starlette = "^0.10.3"
+yarl = "^1.8.1"
+sse-starlette = "^1.1.6"
 python-dotenv = "^0.20.0"
-robonomics-interface = "^1.1.0"
+robonomics-interface = "^1.2.2"
 ecs-logging = "^2.0.0"
 aioprometheus = "^22.5.0"
 
 [tool.poetry.dev-dependencies]
-mypy = "^0.960"
-flake8 = "^4.0.1"
-types-PyYAML = "^6.0.10"
-types-requests = "^2.28.3"
-pytest = "^7.1.2"
+mypy = "^0.971"
+flake8 = "^5.0.4"
+types-PyYAML = "^6.0.11"
+types-requests = "^2.28.9"
+pytest = "^7.1.3"
 vulture = "^2.5"
 pre-commit = "^2.20.0"
 pytest-cov = "^3.0.0"
-ssort = "^0.11.5"
+ssort = "^0.11.6"
 black = {extras = ["d"], version = "^22.6.0"}
 flake8-new-union-types = "^0.4.1"
 flake8-pep585 = "^0.1.5.1"
-pyupgrade = "^2.37.2"
+pyupgrade = "^2.37.3"
 isort = "^5.10.1"
-poetryup = "^0.7.2"
+poetryup = "^0.10.0"
 
 [tool.pytest.ini_options]
 markers = ["short_url"]
diff --git a/src/_workbench_router.py b/src/_workbench_router.py
@@ -152,6 +152,7 @@ def get_schema_list_entry(schema: mdl.ProductionSchema) -> mdl.SchemaListEntry:
         for schema in sorted(all_schemas.values(), key=lambda s: bool(s.is_composite), reverse=True)
         if schema.schema_id not in handled_schemas
     ]
+    available_schemas.sort(key=lambda le: len(le.schema_name))
 
     return mdl.SchemasList(
         status_code=status.HTTP_200_OK,
diff --git a/src/app.py b/src/app.py
@@ -1,3 +1,5 @@
+import os
+
 import uvicorn
 from aioprometheus.asgi.middleware import MetricsMiddleware
 from aioprometheus.asgi.starlette import metrics
@@ -45,6 +47,8 @@
 def startup_event() -> None:
     check_service_connectivity()
     MongoDbWrapper()
+    app_version = os.getenv("VERSION", "Unknown")
+    logger.info(f"Runtime app version: {app_version}")
 
 
 @app.on_event("shutdown")
diff --git a/src/feecc_workbench/Unit.py b/src/feecc_workbench/Unit.py
@@ -11,6 +11,7 @@
 from ._Barcode import Barcode
 from .Employee import Employee
 from .Messenger import messenger
+from .metrics import metrics
 from .models import ProductionSchema
 from .ProductionStage import ProductionStage
 from .Types import AdditionalInfo
@@ -230,5 +231,6 @@ async def end_operation(
                 f"Unit has no more pending production stages. Unit status changed: {prev_status.value} -> "
                 f"{self.status.value}"
             )
+            metrics.register_complete_unit(None, self)
 
         self.employee = None
diff --git a/src/feecc_workbench/WorkBench.py b/src/feecc_workbench/WorkBench.py
@@ -13,6 +13,7 @@
 from .exceptions import StateForbiddenError
 from .ipfs import publish_file
 from .Messenger import messenger
+from .metrics import metrics
 from .models import ProductionSchema
 from .passport_generator import construct_unit_passport
 from .printer import print_image
@@ -73,14 +74,15 @@ async def create_new_unit(self, schema: ProductionSchema) -> Unit:
                 os.remove(unit.barcode.filename)
 
         await self._database.push_unit(unit)
+        metrics.register_create_unit(self.employee, unit)
 
         return unit
 
     def _validate_state_transition(self, new_state: State) -> None:
         """check if state transition can be performed using the map"""
         if new_state not in STATE_TRANSITION_MAP.get(self.state, []):
             message = f"State transition from {self.state.value} to {new_state.value} is not allowed."
-            messenger.error(f"Переход из состояния {self.state.value} в состояние {new_state.value} невозможен")
+            messenger.error("Недопустимая смена состояния")
             raise StateForbiddenError(message)
 
     def switch_state(self, new_state: State) -> None:
@@ -102,6 +104,7 @@ def log_in(self, employee: Employee) -> None:
         messenger.success(f"Авторизован {employee.position} {employee.name}")
 
         self.switch_state(State.AUTHORIZED_IDLING_STATE)
+        metrics.register_log_in(employee)
 
     @logger.catch(reraise=True, exclude=(StateForbiddenError, AssertionError))
     def log_out(self) -> None:
@@ -115,6 +118,7 @@ def log_out(self) -> None:
         message = f"Employee {self.employee.name} was logged out at the workbench no. {self.number}"
         logger.info(message)
         messenger.success(f"{self.employee.name} вышел из системы")
+        metrics.register_log_out(self.employee)
         self.employee = None
 
         self.switch_state(State.AWAIT_LOGIN_STATE)
@@ -143,11 +147,7 @@ def _get_unit_list(unit_: Unit) -> list[Unit]:
 
         if not (override or unit.status in allowed):
             message = f"Can only assign unit with status: {', '.join(s.value for s in allowed)}. Unit status is {unit.status.value}. Forbidden."
-            messenger.warning(
-                f"На стол могут быть помещены изделия со статусами:"
-                f" {', '.join(s.value.upper() for s in allowed)}."
-                f" Статус изделия: {unit.status.value.upper()}. Отказано."
-            )
+            messenger.warning("Сборка изделия уже была завершена, пасспорт выпущен. Отказано.")
             raise AssertionError(message)
 
         self.unit = unit
@@ -267,6 +267,7 @@ async def end_operation(self, additional_info: AdditionalInfo | None = None, pre
         await self._database.push_unit(self.unit, include_components=False)
 
         self.switch_state(State.UNIT_ASSIGNED_IDLING_STATE)
+        metrics.register_complete_operation(self.employee, self.unit)
 
     @logger.catch(reraise=True, exclude=(StateForbiddenError, AssertionError))
     async def upload_unit_passport(self) -> None:
@@ -338,6 +339,7 @@ async def _bg_generate_short_url(url: str, unit_internal_id: str) -> None:
                 asyncio.create_task(post_to_datalog(cid, self.unit.internal_id))
 
         await self._database.push_unit(self.unit)
+        metrics.register_generate_passport(self.employee, self.unit)
 
     async def shutdown(self) -> None:
         logger.info("Workbench shutdown sequence initiated")
diff --git a/src/feecc_workbench/config.py b/src/feecc_workbench/config.py
@@ -1,5 +1,6 @@
 import os
 import sys
+from collections.abc import Iterable
 
 import environ
 from dotenv import load_dotenv
@@ -79,10 +80,29 @@ class HidDevicesNames:
     hid_devices = environ.group(HidDevicesNames)
 
 
+def export_docker_secrets(secret_names: Iterable[str]) -> None:
+    """
+    Export all the requested docker secrets into the environment
+
+    A secret and the corresponding variable are expected to have the same name,
+    but uppercase for the variable and lowercase for the secret file.
+    """
+    for secret in secret_names:
+        secret_path = f"/run/secrets/{secret.lower()}"
+        if not os.path.exists(secret_path):
+            continue
+        with open(secret_path) as f:
+            content = f.read()
+        os.environ[secret.upper()] = content
+        logger.debug(f"Loaded up {secret.upper()} secret from Docker secrets")
+
+
 if __name__ == "__main__":
     print(environ.generate_help(AppConfig))
 
 try:
+    docker_secrets = ["mongodb_uri", "robonomics_account_seed", "yourls_username", "yourls_password"]
+    export_docker_secrets(docker_secrets)
     CONFIG = environ.to_config(AppConfig)
 except environ.MissingEnvValueError as e:
     logger.critical(f"Missing required environment variable '{e}'. Exiting.")
diff --git a/src/feecc_workbench/exceptions.py b/src/feecc_workbench/exceptions.py
@@ -1,16 +1,22 @@
+from copy import copy
 from typing import Any
 
 from .metrics import Metrics
 
 
 class TrackedException(Exception):
+    """An exception that increments Prometheus metric counter for itself"""
+
     _labels: dict[str, str] = {}
 
     def __init__(self, *args: Any) -> None:
+        labels = copy(self._labels)
+        if args:
+            labels["message"] = args[0]
         Metrics().register(
             name=self.__class__.__name__,
             description=self.__class__.__doc__,
-            labels=self._labels,
+            labels=labels,
         )
         super().__init__(*args)
 
@@ -31,3 +37,9 @@ class StateForbiddenError(TrackedException):
     """Raised when state transition is forbidden"""
 
     pass
+
+
+class RobonomicsError(TrackedException):
+    """Raised when Robonmics transactions fail"""
+
+    pass
diff --git a/src/feecc_workbench/metrics.py b/src/feecc_workbench/metrics.py
@@ -1,13 +1,25 @@
+from __future__ import annotations
+
+import os
 import re
+from typing import TYPE_CHECKING
 
-from aioprometheus.collectors import Counter
+from aioprometheus.collectors import Summary
 
+from .Employee import Employee
 from .Singleton import SingletonMeta
+from .utils import export_version
+
+if TYPE_CHECKING:
+    from .Unit import Unit
 
 
 class Metrics(metaclass=SingletonMeta):
     def __init__(self) -> None:
-        self._metrics: dict[str, Counter] = {}
+        self._metrics: dict[str, Summary] = {}
+        export_version()
+        app_version = Summary(name="app_version", doc="Runtime application version")
+        app_version.observe(labels={"app_version": os.getenv("VERSION", "Unknown")}, value=1)
 
     @staticmethod
     def _transform(text: str) -> str:
@@ -16,12 +28,71 @@ def _transform(text: str) -> str:
 
     def _create(self, name: str, description: str) -> None:
         """Create metric"""
-        self._metrics[name] = Counter(name=self._transform(name), doc=description)
+        self._metrics[name] = Summary(name=self._transform(name), doc=description)
 
     def register(self, name: str, description: str | None, labels: dict[str, str] | None = None) -> None:
         """Register metric event"""
         if labels is None:
             labels = {}
         if name not in self._metrics:
             self._create(name=name, description=description or "")
-        self._metrics[name].inc(labels)
+        self._metrics[name].observe(labels=labels, value=1)
+
+    def register_log_in(self, employee: Employee | None) -> None:
+        """Register log_in event"""
+        labels = {
+            "event_type": "log_in",
+            "employee_name": employee.name if employee else "Unknown",
+        }
+        self.register(name="production_metrics", description=None, labels=labels)
+
+    def register_log_out(self, employee: Employee | None) -> None:
+        """Register log_out event"""
+        labels = {
+            "event_type": "log_out",
+            "employee_name": employee.name if employee else "Unknown",
+        }
+        self.register(name="production_metrics", description=None, labels=labels)
+
+    def register_create_unit(self, employee: Employee | None, unit: Unit) -> None:
+        """Register create_unit event"""
+        labels = {
+            "event_type": "create_unit",
+            "employee_name": employee.name if employee else "Unknown",
+            "unit_id": unit.internal_id,
+            "unit_type": unit.schema.unit_name,
+        }
+        self.register(name="production_metrics", description=None, labels=labels)
+
+    def register_complete_unit(self, employee: Employee | None, unit: Unit) -> None:
+        """Register complete_unit event"""
+        labels = {
+            "event_type": "complete_unit",
+            "employee_name": employee.name if employee else "Unknown",
+            "unit_id": unit.internal_id,
+            "unit_type": unit.schema.unit_name,
+        }
+        self.register(name="production_metrics", description=None, labels=labels)
+
+    def register_complete_operation(self, employee: Employee | None, unit: Unit) -> None:
+        """Register complete_operation event"""
+        labels = {
+            "event_type": "complete_operation",
+            "employee_name": employee.name if employee else "Unknown",
+            "unit_id": unit.internal_id,
+            "unit_type": unit.schema.unit_name,
+        }
+        self.register(name="production_metrics", description=None, labels=labels)
+
+    def register_generate_passport(self, employee: Employee | None, unit: Unit) -> None:
+        """Register generate_passport event"""
+        labels = {
+            "event_type": "generate_passport",
+            "employee_name": employee.name if employee else "Unknown",
+            "unit_id": unit.internal_id,
+            "unit_type": unit.schema.unit_name,
+        }
+        self.register(name="production_metrics", description=None, labels=labels)
+
+
+metrics = Metrics()
diff --git a/src/feecc_workbench/robonomics.py b/src/feecc_workbench/robonomics.py
diff --git a/src/feecc_workbench/utils.py b/src/feecc_workbench/utils.py

Original file line number	Diff line number	Diff line change
`@@ -152,6 +152,7 @@ def get_schema_list_entry(schema: mdl.ProductionSchema) -> mdl.SchemaListEntry:`
`152`	`152`	`for schema in sorted(all_schemas.values(), key=lambda s: bool(s.is_composite), reverse=True)`
`153`	`153`	`if schema.schema_id not in handled_schemas`
`154`	`154`	`]`
	`155`	`+ available_schemas.sort(key=lambda le: len(le.schema_name))`
`155`	`156`
`156`	`157`	`return mdl.SchemasList(`
`157`	`158`	`status_code=status.HTTP_200_OK,`