From 0306b222f8e25768b0a934ae4d49e7f47e9be177 Mon Sep 17 00:00:00 2001
From: Waleed Akbar <wakbar@cttc.es>
Date: Mon, 30 Sep 2024 16:14:32 +0000
Subject: [PATCH] Changes to Address Minor Issues in Monitoring Services
Deployment and Build:
- Improved Kafka scripts.
- Updated manifests to handle ports for backend and frontend.
- Changed the return type of a method in settings to `int`.
- Added DB deployment in the service's main file.
---
deploy/kafka.sh | 14 +-
manifests/analyticsservice.yaml | 12 +-
manifests/servicemonitors.yaml | 153 +++++++++++++++++++++
manifests/telemetryservice.yaml | 12 +-
src/analytics/backend/service/__main__.py | 2 +-
src/analytics/frontend/service/__main__.py | 14 +-
src/analytics/tests/test_analytics_db.py | 2 +-
src/common/Settings.py | 10 +-
src/common/tools/kafka/Variables.py | 10 +-
src/kpi_manager/service/__main__.py | 13 +-
src/kpi_manager/tests/test_kpi_db.py | 2 +-
src/telemetry/backend/service/__main__.py | 2 +-
src/telemetry/frontend/service/__main__.py | 12 +-
src/telemetry/tests/test_telemetryDB.py | 2 +-
14 files changed, 230 insertions(+), 30 deletions(-)
diff --git a/deploy/kafka.sh b/deploy/kafka.sh
index 0483bce15..4cbcdb701 100755
--- a/deploy/kafka.sh
+++ b/deploy/kafka.sh
@@ -47,10 +47,10 @@ function kafka_deploy() {
cp "${KFK_MANIFESTS_PATH}/${KFK_MANIFEST}" "${TMP_MANIFESTS_FOLDER}/${KFK_MANIFEST}"
# echo "Apache Kafka Namespace"
- echo ">>> Delete Apache Kafka Namespace"
+ echo "Delete Apache Kafka Namespace"
kubectl delete namespace ${KFK_NAMESPACE} --ignore-not-found
- echo ">>> Create Apache Kafka Namespace"
+ echo "Create Apache Kafka Namespace"
kubectl create namespace ${KFK_NAMESPACE}
# echo ">>> Deplying Apache Kafka Zookeeper"
@@ -76,15 +76,15 @@ function kafka_deploy() {
# fi
}
-echo "Apache Kafka"
-echo ">>> Checking if Apache Kafka is deployed ... "
+echo ">>> Apache Kafka"
+echo "Checking if Apache Kafka is deployed ... "
if [ "$KFK_REDEPLOY" == "YES" ]; then
- echo ">>> Redeploying kafka namespace"
+ echo "Redeploying kafka namespace"
kafka_deploy
elif kubectl get namespace "${KFK_NAMESPACE}" &> /dev/null; then
- echo ">>> Apache Kafka already present; skipping step."
+ echo "Apache Kafka already present; skipping step."
else
- echo ">>> Kafka namespace doesn't exists. Deploying kafka namespace"
+ echo "Kafka namespace doesn't exists. Deploying kafka namespace"
kafka_deploy
fi
echo
diff --git a/manifests/analyticsservice.yaml b/manifests/analyticsservice.yaml
index 7340dff5f..e15214dbd 100644
--- a/manifests/analyticsservice.yaml
+++ b/manifests/analyticsservice.yaml
@@ -39,6 +39,8 @@ spec:
value: "INFO"
- name: CRDB_DATABASE
value: "tfs_analytics"
+ - name: METRICS_PORT
+ value: "9192"
envFrom:
- secretRef:
name: crdb-analytics
@@ -62,10 +64,12 @@ spec:
imagePullPolicy: Always
ports:
- containerPort: 30090
- - containerPort: 9192
+ - containerPort: 9193
env:
- name: LOG_LEVEL
value: "INFO"
+ - name: METRICS_PORT
+ value: "9193"
envFrom:
- secretRef:
name: kfk-kpi-data
@@ -102,10 +106,14 @@ spec:
protocol: TCP
port: 30090
targetPort: 30090
- - name: metrics
+ - name: metrics-frontend
protocol: TCP
port: 9192
targetPort: 9192
+ - name: metrics-backend
+ protocol: TCP
+ port: 9193
+ targetPort: 9193
---
apiVersion: autoscaling/v2
kind: HorizontalPodAutoscaler
diff --git a/manifests/servicemonitors.yaml b/manifests/servicemonitors.yaml
index 716c1c689..8a8fe6f39 100644
--- a/manifests/servicemonitors.yaml
+++ b/manifests/servicemonitors.yaml
@@ -475,3 +475,156 @@ spec:
any: false
matchNames:
- tfs # namespace where the app is running
+---
+apiVersion: monitoring.coreos.com/v1
+kind: ServiceMonitor
+metadata:
+ namespace: monitoring # namespace where prometheus is running
+ name: tfs-analyticsservice-metric
+ labels:
+ app: analyticsservice
+ #release: prometheus
+ #release: prom # name of the release
+ # ( VERY IMPORTANT: You need to know the correct release name by viewing
+ # the servicemonitor of Prometheus itself: Without the correct name,
+ # Prometheus cannot identify the metrics of the Flask app as the target.)
+spec:
+ selector:
+ matchLabels:
+ # Target app service
+ #namespace: tfs
+ app: analyticsservice # same as above
+ #release: prometheus # same as above
+ endpoints:
+ - port: metrics-frontend # named port in target app
+ scheme: http
+ path: /metrics # path to scrape
+ interval: 5s # scrape interval
+ - port: metrics-backend # named port in target app
+ scheme: http
+ path: /metrics # path to scrape
+ interval: 5s # scrape interval
+ namespaceSelector:
+ any: false
+ matchNames:
+ - tfs # namespace where the app is running
+---
+apiVersion: monitoring.coreos.com/v1
+kind: ServiceMonitor
+metadata:
+ namespace: monitoring # namespace where prometheus is running
+ name: tfs-telemetryservice-metric
+ labels:
+ app: telemetryservice
+ #release: prometheus
+ #release: prom # name of the release
+ # ( VERY IMPORTANT: You need to know the correct release name by viewing
+ # the servicemonitor of Prometheus itself: Without the correct name,
+ # Prometheus cannot identify the metrics of the Flask app as the target.)
+spec:
+ selector:
+ matchLabels:
+ # Target app service
+ #namespace: tfs
+ app: telemetryservice # same as above
+ #release: prometheus # same as above
+ endpoints:
+ - port: metrics-frontend # named port in target app
+ scheme: http
+ path: /metrics # path to scrape
+ interval: 5s # scrape interval
+ - port: metrics-backend # named port in target app
+ scheme: http
+ path: /metrics # path to scrape
+ interval: 5s # scrape interval
+ namespaceSelector:
+ any: false
+ matchNames:
+ - tfs # namespace where the app is running
+---
+apiVersion: monitoring.coreos.com/v1
+kind: ServiceMonitor
+metadata:
+ namespace: monitoring # namespace where prometheus is running
+ name: tfs-kpi-managerservice-metric
+ labels:
+ app: kpi-managerservice
+ #release: prometheus
+ #release: prom # name of the release
+ # ( VERY IMPORTANT: You need to know the correct release name by viewing
+ # the servicemonitor of Prometheus itself: Without the correct name,
+ # Prometheus cannot identify the metrics of the Flask app as the target.)
+spec:
+ selector:
+ matchLabels:
+ # Target app service
+ #namespace: tfs
+ app: kpi-managerservice # same as above
+ #release: prometheus # same as above
+ endpoints:
+ - port: metrics # named port in target app
+ scheme: http
+ path: /metrics # path to scrape
+ interval: 5s # scrape interval
+ namespaceSelector:
+ any: false
+ matchNames:
+ - tfs # namespace where the app is running
+---
+apiVersion: monitoring.coreos.com/v1
+kind: ServiceMonitor
+metadata:
+ namespace: monitoring # namespace where prometheus is running
+ name: tfs-kpi_value_apiservice-metric
+ labels:
+ app: kpi_value_apiservice
+ #release: prometheus
+ #release: prom # name of the release
+ # ( VERY IMPORTANT: You need to know the correct release name by viewing
+ # the servicemonitor of Prometheus itself: Without the correct name,
+ # Prometheus cannot identify the metrics of the Flask app as the target.)
+spec:
+ selector:
+ matchLabels:
+ # Target app service
+ #namespace: tfs
+ app: kpi_value_apiservice # same as above
+ #release: prometheus # same as above
+ endpoints:
+ - port: metrics # named port in target app
+ scheme: http
+ path: /metrics # path to scrape
+ interval: 5s # scrape interval
+ namespaceSelector:
+ any: false
+ matchNames:
+ - tfs # namespace where the app is running
+---
+apiVersion: monitoring.coreos.com/v1
+kind: ServiceMonitor
+metadata:
+ namespace: monitoring # namespace where prometheus is running
+ name: tfs-kpi_value_writerservice-metric
+ labels:
+ app: kpi_value_writerservice
+ #release: prometheus
+ #release: prom # name of the release
+ # ( VERY IMPORTANT: You need to know the correct release name by viewing
+ # the servicemonitor of Prometheus itself: Without the correct name,
+ # Prometheus cannot identify the metrics of the Flask app as the target.)
+spec:
+ selector:
+ matchLabels:
+ # Target app service
+ #namespace: tfs
+ app: kpi_value_writerservice # same as above
+ #release: prometheus # same as above
+ endpoints:
+ - port: metrics # named port in target app
+ scheme: http
+ path: /metrics # path to scrape
+ interval: 5s # scrape interval
+ namespaceSelector:
+ any: false
+ matchNames:
+ - tfs # namespace where the app is running
diff --git a/manifests/telemetryservice.yaml b/manifests/telemetryservice.yaml
index 7c781bb3d..9c979713d 100644
--- a/manifests/telemetryservice.yaml
+++ b/manifests/telemetryservice.yaml
@@ -39,6 +39,8 @@ spec:
value: "INFO"
- name: CRDB_DATABASE
value: "tfs_kpi"
+ - name: METRICS_PORT
+ value: "9192"
envFrom:
- secretRef:
name: crdb-telemetry
@@ -62,10 +64,12 @@ spec:
imagePullPolicy: Always
ports:
- containerPort: 30060
- - containerPort: 9192
+ - containerPort: 9193
env:
- name: LOG_LEVEL
value: "INFO"
+ - name: METRICS_PORT
+ value: "9193"
envFrom:
- secretRef:
name: kfk-kpi-data
@@ -102,10 +106,14 @@ spec:
protocol: TCP
port: 30060
targetPort: 30060
- - name: metrics
+ - name: metrics-frontend
protocol: TCP
port: 9192
targetPort: 9192
+ - name: metrics-backend
+ protocol: TCP
+ port: 9193
+ targetPort: 9193
---
apiVersion: autoscaling/v2
kind: HorizontalPodAutoscaler
diff --git a/src/analytics/backend/service/__main__.py b/src/analytics/backend/service/__main__.py
index 3c4c36b7c..9b1941e20 100644
--- a/src/analytics/backend/service/__main__.py
+++ b/src/analytics/backend/service/__main__.py
@@ -37,7 +37,7 @@ def main():
LOGGER.info('Starting...')
# Start metrics server
- metrics_port = get_metrics_port()
+ metrics_port = int(get_metrics_port())
start_http_server(metrics_port)
grpc_service = AnalyticsBackendService()
diff --git a/src/analytics/frontend/service/__main__.py b/src/analytics/frontend/service/__main__.py
index 6c331844f..bf94eb5e6 100644
--- a/src/analytics/frontend/service/__main__.py
+++ b/src/analytics/frontend/service/__main__.py
@@ -16,9 +16,14 @@ import logging, signal, sys, threading
from prometheus_client import start_http_server
from common.Settings import get_log_level, get_metrics_port
from .AnalyticsFrontendService import AnalyticsFrontendService
+from analytics.database.AnalyzerModel import Analyzer as Model
+from common.tools.database.GenericDatabase import Database
+from common.Settings import get_setting
+
+DB_NAME = get_setting('CRDB_DATABASE', default=None)
terminate = threading.Event()
-LOGGER = None
+LOGGER = None
def signal_handler(signal, frame): # pylint: disable=redefined-outer-name
LOGGER.warning('Terminate signal received')
@@ -36,8 +41,13 @@ def main():
LOGGER.info('Starting...')
+ # To create DB
+ kpiDBobj = Database(DB_NAME, Model)
+ kpiDBobj.create_database()
+ kpiDBobj.create_tables()
+
# Start metrics server
- metrics_port = get_metrics_port()
+ metrics_port = int(get_metrics_port())
start_http_server(metrics_port)
grpc_service = AnalyticsFrontendService()
diff --git a/src/analytics/tests/test_analytics_db.py b/src/analytics/tests/test_analytics_db.py
index f944fc0b5..58e7d0167 100644
--- a/src/analytics/tests/test_analytics_db.py
+++ b/src/analytics/tests/test_analytics_db.py
@@ -14,7 +14,7 @@
import logging
-from analytics.database.new_Analyzer_DB import AnalyzerDB
+from analytics.database.Analyzer_DB import AnalyzerDB
LOGGER = logging.getLogger(__name__)
diff --git a/src/common/Settings.py b/src/common/Settings.py
index eaeb363ad..936c0387b 100644
--- a/src/common/Settings.py
+++ b/src/common/Settings.py
@@ -79,12 +79,12 @@ def get_service_host(service_name : ServiceNameEnum):
def get_service_port_grpc(service_name : ServiceNameEnum):
envvar_name = get_env_var_name(service_name, ENVVAR_SUFIX_SERVICE_PORT_GRPC)
default_value = DEFAULT_SERVICE_GRPC_PORTS.get(service_name.value)
- return get_setting(envvar_name, default=default_value)
+ return int(get_setting(envvar_name, default=default_value))
def get_service_port_http(service_name : ServiceNameEnum):
envvar_name = get_env_var_name(service_name, ENVVAR_SUFIX_SERVICE_PORT_HTTP)
default_value = DEFAULT_SERVICE_HTTP_PORTS.get(service_name.value)
- return get_setting(envvar_name, default=default_value)
+ return int(get_setting(envvar_name, default=default_value))
def get_service_baseurl_http(service_name : ServiceNameEnum):
envvar_name = get_env_var_name(service_name, ENVVAR_SUFIX_SERVICE_BASEURL_HTTP)
@@ -95,16 +95,16 @@ def get_log_level():
return get_setting(ENVVAR_LOG_LEVEL, default=DEFAULT_LOG_LEVEL)
def get_metrics_port():
- return get_setting(ENVVAR_METRICS_PORT, default=DEFAULT_METRICS_PORT)
+ return int(get_setting(ENVVAR_METRICS_PORT, default=DEFAULT_METRICS_PORT))
def get_grpc_bind_address():
return get_setting(ENVVAR_GRPC_BIND_ADDRESS, default=DEFAULT_GRPC_BIND_ADDRESS)
def get_grpc_max_workers():
- return get_setting(ENVVAR_GRPC_MAX_WORKERS, default=DEFAULT_GRPC_MAX_WORKERS)
+ return int(get_setting(ENVVAR_GRPC_MAX_WORKERS, default=DEFAULT_GRPC_MAX_WORKERS))
def get_grpc_grace_period():
- return get_setting(ENVVAR_GRPC_GRACE_PERIOD, default=DEFAULT_GRPC_GRACE_PERIOD)
+ return int(get_setting(ENVVAR_GRPC_GRACE_PERIOD, default=DEFAULT_GRPC_GRACE_PERIOD))
def get_http_bind_address():
return get_setting(ENVVAR_HTTP_BIND_ADDRESS, default=DEFAULT_HTTP_BIND_ADDRESS)
diff --git a/src/common/tools/kafka/Variables.py b/src/common/tools/kafka/Variables.py
index fc43c3151..73b633e23 100644
--- a/src/common/tools/kafka/Variables.py
+++ b/src/common/tools/kafka/Variables.py
@@ -25,11 +25,11 @@ class KafkaConfig(Enum):
@staticmethod
def get_kafka_address() -> str:
- # kafka_server_address = get_setting('KFK_SERVER_ADDRESS', default=None)
- # if kafka_server_address is None:
- KFK_NAMESPACE = get_setting('KFK_NAMESPACE')
- KFK_PORT = get_setting('KFK_SERVER_PORT')
- kafka_server_address = KFK_SERVER_ADDRESS_TEMPLATE.format(KFK_NAMESPACE, KFK_PORT)
+ kafka_server_address = get_setting('KFK_SERVER_ADDRESS', default=None)
+ if kafka_server_address is None:
+ KFK_NAMESPACE = get_setting('KFK_NAMESPACE')
+ KFK_PORT = get_setting('KFK_SERVER_PORT')
+ kafka_server_address = KFK_SERVER_ADDRESS_TEMPLATE.format(KFK_NAMESPACE, KFK_PORT)
return kafka_server_address
@staticmethod
diff --git a/src/kpi_manager/service/__main__.py b/src/kpi_manager/service/__main__.py
index 244d5afa3..6a3d078e2 100644
--- a/src/kpi_manager/service/__main__.py
+++ b/src/kpi_manager/service/__main__.py
@@ -16,8 +16,14 @@ import logging, signal, sys, threading
from common.Settings import get_log_level
from .KpiManagerService import KpiManagerService
+from kpi_manager.database.KpiModel import Kpi as Model
+from common.tools.database.GenericDatabase import Database
+from common.Settings import get_setting
+
+
+DB_NAME = get_setting('CRDB_DATABASE', default=None)
terminate = threading.Event()
-LOGGER = None
+LOGGER = None
def signal_handler(signal, frame): # pylint: disable=redefined-outer-name
LOGGER.warning('Terminate signal received')
@@ -35,6 +41,11 @@ def main():
LOGGER.debug('Starting...')
+ # To create DB
+ kpiDBobj = Database(DB_NAME, Model)
+ kpiDBobj.create_database()
+ kpiDBobj.create_tables()
+
grpc_service = KpiManagerService()
grpc_service.start()
diff --git a/src/kpi_manager/tests/test_kpi_db.py b/src/kpi_manager/tests/test_kpi_db.py
index b9c3b2d7c..b1513a83f 100644
--- a/src/kpi_manager/tests/test_kpi_db.py
+++ b/src/kpi_manager/tests/test_kpi_db.py
@@ -18,7 +18,7 @@ import logging
from common.proto.kpi_manager_pb2 import KpiDescriptorList
from .test_messages import create_kpi_filter_request
from kpi_manager.database.KpiModel import Kpi as KpiModel
-from kpi_manager.database.new_KpiDB import KpiDB
+from kpi_manager.database.KpiDB import KpiDB
# from common.tools.database.GenericDatabase import Database
LOGGER = logging.getLogger(__name__)
diff --git a/src/telemetry/backend/service/__main__.py b/src/telemetry/backend/service/__main__.py
index 9ec9e191f..8af209e2c 100644
--- a/src/telemetry/backend/service/__main__.py
+++ b/src/telemetry/backend/service/__main__.py
@@ -37,7 +37,7 @@ def main():
LOGGER.info('Starting...')
# Start metrics server
- metrics_port = get_metrics_port()
+ metrics_port = int(get_metrics_port())
start_http_server(metrics_port)
grpc_service = TelemetryBackendService()
diff --git a/src/telemetry/frontend/service/__main__.py b/src/telemetry/frontend/service/__main__.py
index 2a6c5dbcf..60b96646a 100644
--- a/src/telemetry/frontend/service/__main__.py
+++ b/src/telemetry/frontend/service/__main__.py
@@ -16,7 +16,12 @@ import logging, signal, sys, threading
from prometheus_client import start_http_server
from common.Settings import get_log_level, get_metrics_port
from .TelemetryFrontendService import TelemetryFrontendService
+from telemetry.database.TelemetryModel import Collector as Model
+from common.tools.database.GenericDatabase import Database
+from common.Settings import get_setting
+
+DB_NAME = get_setting('CRDB_DATABASE', default=None)
terminate = threading.Event()
LOGGER = None
@@ -36,8 +41,13 @@ def main():
LOGGER.info('Starting...')
+ # To create DB
+ kpiDBobj = Database(DB_NAME, Model)
+ kpiDBobj.create_database()
+ kpiDBobj.create_tables()
+
# Start metrics server
- metrics_port = get_metrics_port()
+ metrics_port = int(get_metrics_port())
start_http_server(metrics_port)
grpc_service = TelemetryFrontendService()
diff --git a/src/telemetry/tests/test_telemetryDB.py b/src/telemetry/tests/test_telemetryDB.py
index 707e6b5b2..0d5bd0584 100644
--- a/src/telemetry/tests/test_telemetryDB.py
+++ b/src/telemetry/tests/test_telemetryDB.py
@@ -14,7 +14,7 @@
import logging
-from telemetry.database.new_Telemetry_DB import TelemetryDB
+from telemetry.database.Telemetry_DB import TelemetryDB
LOGGER = logging.getLogger(__name__)
--
GitLab