new request logging

.gitignore

@@ -117,6 +117,9 @@ wrappers/python/fbs/
 examples/models/onnx_resnet50/cpu_codegen/Function_0_codegen.cpp
 examples/models/onnx_resnet50/resnet.onnx
 
+#logging example
+examples/centralised-logging/request-logging/istio-1.1.6
+
 #openapi
 engine/src/main/resources/static/seldon.json
 api-frontend/src/main/resources/static/seldon.json

doc/source/analytics/analytics.md

@@ -6,8 +6,8 @@ The metrics are:
 
 **Prediction Requests**
 
- * ```seldon_api_executor_server_requests_duration_seconds_(bucket,count,sum) ``` : Requests to the service orchestrator from an ingress, e.g. API gateway or Ambassador
- * ```seldon_api_executor_client_requests_duration_seconds_(bucket,count,sum) ``` : Requests from the service orchestrator to a component, e.g., a model
+ * ```seldon_api_executor_server_requests_seconds_(bucket,count,sum) ``` : Requests to the service orchestrator from an ingress, e.g. API gateway or Ambassador
+ * ```seldon_api_executor_client_requests_seconds_(bucket,count,sum) ``` : Requests from the service orchestrator to a component, e.g., a model
 
 Each metric has the following key value pairs for further filtering which will be taken from the SeldonDeployment custom resource that is running:
 

examples/centralised-logging/README.md

@@ -4,13 +4,29 @@
 
 Here we will set up EFK (elasticsearch, fluentd/fluentbit, kibana) as a stack to gather logs from SeldonDeployments and make them searchable.
 
-This demo is aimed at minikube.
+This demo is aimed at KIND or minikube but can also work with a cloud provider. Uses helm v3.
 
 Alternatives are available and if you are running in cloud then you can consider a managed service from your cloud provider.
 
 If you just want to bootstrap a full logging and request tracking setup for minikube, run ./full-setup.sh. That includes the [request logging setup](./request-logging/README.md)
 
-## Setup
+## Setup Elastic - KIND
+
+Start cluster
+
+```
+kind create cluster --config kind_config.yaml --image kindest/node:v1.15.6
+```
+
+Install elastic with KIND config:
+
+```
+kubectl create namespace logs
+kubectl apply -f https://raw.githubusercontent.com/rancher/local-path-provisioner/master/deploy/local-path-storage.yaml
+helm install elasticsearch elasticsearch --version 7.5.2 --namespace=logs -f elastic-kind.yaml --repo https://helm.elastic.co
+```
+
+## Setup Elastic - Minikube
 
 Start Minikube with flags as shown:
 
@@ -22,11 +38,10 @@ Install elasticsearch with minikube configuration:
 
 ```
 kubectl create namespace logs
+helm install elasticsearch elasticsearch --version 7.5.2 --namespace=logs -f elastic-minikube.yaml --repo https://helm.elastic.co
 ```
 
-```
-helm install elasticsearch elasticsearch --version 7.1.1 --namespace=logs -f elastic-minikube.yaml --repo https://helm.elastic.co
-```
+## Fluentd and Kibana
 
 Then fluentd as a collection agent (chosen in preference to fluentbit - see notes at end):
 
@@ -37,7 +52,7 @@ helm install fluentd fluentd-elasticsearch --namespace=logs -f fluentd-values.ya
 And kibana UI:
 
 ```
-helm install kibana kibana --version 7.1.1 --namespace=logs --set service.type=NodePort --repo https://helm.elastic.co
+helm install kibana kibana --version 7.5.2 --namespace=logs --set service.type=NodePort --repo https://helm.elastic.co
 ```
 
 ## Generating Logging
@@ -57,15 +72,16 @@ Check that it now recognises the seldon CRD by running `kubectl get sdep`.
 Now a model:
 
 ```
-helm install seldon-single-model ../../helm-charts/seldon-single-model/ --set engine.env.LOG_MESSAGES_EXTERNALLY="false"
+helm install seldon-single-model ../../helm-charts/seldon-single-model/
 ```
 
-And the loadtester:
+And the loadtester (first line is only needed for KIND):
 
 ```
+kubectl label nodes kind-worker role=locust --overwrite
 kubectl label nodes $(kubectl get nodes -o jsonpath='{.items[0].metadata.name}') role=locust --overwrite
 
-helm install seldon-core-loadtesting ../../helm-charts/seldon-core-loadtesting/ --set locust.host=http://seldon-single-model-seldon-single-model:8000 --set oauth.enabled=false --set oauth.key=oauth-key --set oauth.secret=oauth-secret --set locust.hatchRate=1 --set locust.clients=1 --set loadtest.sendFeedback=0 --set locust.minWait=0 --set locust.maxWait=0 --set replicaCount=1
+helm install seldon-core-loadtesting ../../helm-charts/seldon-core-loadtesting/ --set locust.host=http://seldon-single-model-seldon-single-model-seldon-single-model:8000 --set oauth.enabled=false --set oauth.key=oauth-key --set oauth.secret=oauth-secret --set locust.hatchRate=1 --set locust.clients=1 --set loadtest.sendFeedback=0 --set locust.minWait=1000 --set locust.maxWait=1000 --set replicaCount=1
 ```
 
 ## Inspecting Logging and Search for Requests
@@ -75,11 +91,17 @@ To find kibana URL
 ```
 echo $(minikube ip)":"$(kubectl get svc kibana-kibana -n logs -o=jsonpath='{.spec.ports[?(@.port==5601)].nodePort}')
 ```
+Or if not on minikube then port-forward to `localhost:5601`:
+```
+kubectl port-forward svc/kibana-kibana -n logs 5601:5601
+```
+
+If you want to check the elastic API with postman then also run `kubectl port-forward svc/elasticsearch-master -n logs 9200:9200`
 
 When Kibana appears for the first time there will be a brief animation while it initializes.
 On the Welcome page click Explore on my own.
 From the top-left or from the `Visualize and Explore Data` panel select the `Discover` item.
-In the form field Index pattern enter logstash-*
+In the form field Index pattern enter *
 It should read "Success!" and Click the `> Next` step button on the right.
 In the next form select timestamp from the dropdown labeled `Time Filter` field name.
 From the bottom-right of the form select `Create index pattern`.
@@ -88,13 +110,11 @@ From the top-left or the home screen's `Visualize and Explore Data` panel, selec
 The log list will appear.
 Refine the list a bit by selecting `log` near the bottom the left-hand Selected fields list.
 When you hover over or click on the word `log`, click the `Add` button to the right of the label.
-You can create a filter using the `Add Filter` button under `Search`. The field can be `kubernetes.labels.seldon-app` and the value can be an 'is' match on `seldon-single-model-seldon-single-model`.
-
-The custom fields in the request bodies may not currently be in the index. If you hover over one in a request you may see `No cached mapping for this field`.
+You can create a filter using the `Add Filter` button under `Search`. The field can be `kubernetes.labels.seldon-app` and the value can be an 'is' match on `seldon-single-model-seldon-single-model-seldon-single-model`.
 
-To add mappings, go to `Management` at the bottom-left and then `Index Patterns`. Hit `Refresh` on the index created earlier. The number of fields should increase and `request.data.names` should be present.
+To add mappings, go to `Management` at the bottom-left and then `Index Patterns`. Hit `Refresh` on the index created earlier. The number of fields should increase.
 
-Now we can go back and add a further filter for `data.names` with the operator `exists`. We can add further filters if we want, such as the presence of a feature name or the presence of a feature value.
+Now we can go back and add further filters if we want.
 
 ![picture](./kibana-custom-search.png)
 

examples/centralised-logging/elastic-kind.yaml

@@ -0,0 +1,45 @@
+---
+# Permit co-located instances for solitary minikube virtual machines.
+antiAffinity: "soft"
+
+# Shrink default JVM heap.
+esJavaOpts: "-Xmx256m -Xms256m"
+
+podAnnotations:
+  fluentbit.io/exclude: "true"
+
+replicas: 1
+
+# Allocate smaller chunks of memory per pod.
+resources:
+  requests:
+    cpu: "200m"
+    memory: "512M"
+  limits:
+    cpu: "1500m"
+    memory: "1024M"
+
+# Request smaller persistent volumes.
+volumeClaimTemplate:
+  accessModes: [ "ReadWriteOnce" ]
+  storageClassName: "local-path"
+  resources:
+    requests:
+      storage: 400M
+extraInitContainers: |
+   - name: create
+     image: busybox:1.28
+     command: ['mkdir', '-p', '/usr/share/elasticsearch/data/nodes/']
+     securityContext:
+       runAsUser: 0
+     volumeMounts:
+      - mountPath: /usr/share/elasticsearch/data
+        name: elasticsearch-master
+   - name: file-permissions
+     image: busybox:1.28
+     command: ['chown', '-R', '1000:1000', '/usr/share/elasticsearch/']
+     securityContext:
+        runAsUser: 0
+     volumeMounts:
+      - mountPath: /usr/share/elasticsearch/data
+        name: elasticsearch-master

examples/centralised-logging/full-setup-existing-kubeflow.sh

@@ -9,18 +9,35 @@ set -o xtrace
 
 # Assumes existing cluster with kubeflow's istio gateway
 # Will put services behind kubeflow istio gateway
-brokercrd=$(kubectl get crd inmemorychannels.messaging.knative.dev -o jsonpath='{.metadata.name}') || true
+# First check what parts of knative are present
+autoscaler=$(kubectl get deployment -n knative-serving autoscaler -o jsonpath='{.metadata.name}') || true
+if [[ $autoscaler == 'autoscaler' ]] ; then
+    echo "knative serving already installed"
+else
+   ./request-logging/install_knative.sh
+fi
+
+imc=$(kubectl get deployment -n knative-eventing imc-controller -o jsonpath='{.metadata.name}') || true
 
-if [[ $brokercrd == 'inmemorychannels.messaging.knative.dev' ]] ; then
-    echo "knative already installed"
+if [[ $imc == 'imc-controller' ]] ; then
+    echo "knative eventing already installed"
 else
-   ./kubeflow/knative-setup-existing-istio.sh
+   kubectl apply --selector knative.dev/crd-install=true --filename https://github.com/knative/eventing/releases/download/v0.11.0/eventing.yaml
+   sleep 5
+   kubectl apply --filename https://github.com/knative/eventing/releases/download/v0.11.0/eventing.yaml
+   kubectl apply --filename https://github.com/knative/eventing/releases/download/v0.11.0/in-memory-channel.yaml
 fi
 
+#istio for knative needs to have cluster-local-gateway
+#script installs any missing istio components (but leaves existing ones)
+cd request-logging
+./install_istio.sh
+cd ..
+
 sleep 5
 
 kubectl create namespace seldon-system || echo "namespace seldon-system exists"
-helm upgrade --install seldon-core ../../helm-charts/seldon-core-operator/ --namespace seldon-system --set istio.gateway="kubeflow-gateway.kubeflow.svc.cluster.local" --set istio.enabled="true" --set engine.logMessagesExternally="true" --set certManager.enabled="true"
+helm upgrade --install seldon-core ../../helm-charts/seldon-core-operator/ --namespace seldon-system --set istio.gateway="kubeflow-gateway.kubeflow.svc.cluster.local" --set istio.enabled="true" --set certManager.enabled="true"
 
 kubectl rollout status -n seldon-system deployment/seldon-controller-manager
 
@@ -29,22 +46,28 @@ sleep 5
 helm upgrade --install seldon-core-analytics ../../helm-charts/seldon-core-analytics/ --namespace default -f ./kubeflow/seldon-analytics-kubeflow.yaml
 
 kubectl create namespace logs || echo "namespace logs exists"
-helm upgrade --install elasticsearch elasticsearch --version 7.5.0 --namespace=logs --set service.type=ClusterIP --set antiAffinity="soft" --repo https://helm.elastic.co
+helm upgrade --install elasticsearch elasticsearch --version 7.5.2 --namespace=logs --set service.type=ClusterIP --set antiAffinity="soft" --repo https://helm.elastic.co
 kubectl rollout status statefulset/elasticsearch-master -n logs
 
 helm upgrade --install fluentd fluentd-elasticsearch --namespace=logs -f fluentd-values.yaml --repo https://kiwigrid.github.io
-helm upgrade --install kibana kibana --version 7.5.0 --namespace=logs --set service.type=ClusterIP -f ./kubeflow/kibana-values.yaml --repo https://helm.elastic.co
+helm upgrade --install kibana kibana --version 7.5.2 --namespace=logs --set service.type=ClusterIP -f ./kubeflow/kibana-values.yaml --repo https://helm.elastic.co
 
 kubectl apply -f ./kubeflow/virtualservice-kibana.yaml
 kubectl apply -f ./kubeflow/virtualservice-elasticsearch.yaml
 
 kubectl rollout status deployment/kibana-kibana -n logs
 
+#have to delete logger if existing as otherwise get 'expected exactly one, got both' err if existing resource is v1alpha1
+kubectl delete -f ./request-logging/seldon-request-logger.yaml || true
 kubectl apply -f ./request-logging/seldon-request-logger.yaml
+# remove and recreate broker if already have one to activate eventing
+kubectl delete broker -n default default || true
+kubectl label namespace default knative-eventing-injection- --overwrite=true
 kubectl label namespace default knative-eventing-injection=enabled --overwrite=true
 #sleep 3
 sleep 6
 kubectl -n default get broker default
+
 kubectl apply -f ./request-logging/trigger.yaml
 
 ISTIO_INGRESS=$(kubectl get svc -n istio-system istio-ingressgateway -o jsonpath='{.status.loadBalancer.ingress[0].ip}')

examples/centralised-logging/full-setup.sh

@@ -19,12 +19,12 @@ helm install --name seldon-core-loadtesting ../../helm-charts/seldon-core-loadte
 
 
 kubectl create namespace logs || echo "namespace logs exists"
-helm install --name elasticsearch elasticsearch --version 7.5.0 --namespace=logs -f elastic-minikube.yaml --repo https://helm.elastic.co
+helm install --name elasticsearch elasticsearch --version 7.5.2 --namespace=logs -f elastic-minikube.yaml --repo https://helm.elastic.co
 kubectl rollout status statefulset/elasticsearch-master -n logs
 kubectl patch svc elasticsearch-master -n logs -p '{"spec": {"type": "LoadBalancer"}}'
 
 helm install fluentd-elasticsearch --name fluentd --namespace=logs -f fluentd-values.yaml --repo https://kiwigrid.github.io
-helm install kibana --version 7.5.0 --name=kibana --namespace=logs --set service.type=NodePort --repo https://helm.elastic.co
+helm install kibana --version 7.5.2 --name=kibana --namespace=logs --set service.type=NodePort --repo https://helm.elastic.co
 
 kubectl rollout status deployment/kibana-kibana -n logs
 

examples/centralised-logging/kind_config.yaml

@@ -0,0 +1,32 @@
+kind: Cluster
+apiVersion: kind.sigs.k8s.io/v1alpha3
+nodes:
+- role: control-plane
+- role: worker
+  extraPortMappings:
+  - containerPort: 30080
+    hostPort: 8003
+  - containerPort: 31380
+    hostPort: 8004
+kubeadmConfigPatches:
+- |
+  apiVersion: kubelet.config.k8s.io/v1beta1
+  kind: KubeletConfiguration
+  metadata:
+    name: config
+  kubeReserved:
+      cpu: "300m"
+      memory: "300Mi"
+      ephemeral-storage: "1Gi"
+  kubeReservedCgroup: "/kube-reserved"
+  systemReserved:
+      cpu: "300m"
+      memory: "300Mi"
+      ephemeral-storage: "1Gi"
+  evictionHard:
+      memory.available:  "200Mi"
+      nodefs.available: "10%"
+  featureGates:
+      DynamicKubeletConfig: true
+      RotateKubeletServerCertificate: true
+

examples/centralised-logging/kubeflow/knative-setup-existing-istio.sh

@@ -1,17 +1,3 @@
 #this assumes installing to cloud and istio already installed e.g. with kubeflow
 
-kubectl apply --selector knative.dev/crd-install=true \
---filename https://github.com/knative/serving/releases/download/v0.8.0/serving.yaml \
---filename https://github.com/knative/eventing/releases/download/v0.8.0/release.yaml \
---filename https://github.com/knative/serving/releases/download/v0.8.0/monitoring.yaml
-
-kubectl apply --filename https://github.com/knative/serving/releases/download/v0.8.0/serving.yaml \
---filename https://github.com/knative/eventing/releases/download/v0.8.0/release.yaml \
---filename https://github.com/knative/serving/releases/download/v0.8.0/monitoring.yaml
-
-kubectl label namespace default istio-injection=enabled
-
-kubectl apply -f https://github.com/knative/eventing/releases/download/v0.8.0/eventing.yaml
-
-#kafka if you have a kafka cluster setup already
-#kubectl apply -f https://github.com/knative/eventing/releases/download/v0.8.0/kafka.yaml
+./../request-logging/install_knative.sh

examples/centralised-logging/kubeflow/seldon-analytics-kubeflow.yaml

@@ -1,15 +1,6 @@
-grafana_prom_service_type: ClusterIP
-grafana_prom_admin_password: admin
-grafana_anonymous_auth: true
-grafana:
-  virtualservice:
-    enabled: true
-    #trailing dash important and should be used when accessing
-    prefix: "/grafana/"
-    gateways:
-      - kubeflow-gateway.kubeflow.svc.cluster.local
-  extraEnv:
-    #replace with KF gateway URI
-    GF_SERVER_ROOT_URL: "%(protocol)s://%(domain)s/grafana"
-nodeExporter:
-  port: 9200
+prometheus:
+  nodeExporter:
+    hostNetwork: false
+    service:
+      hostPort: 9200
+      servicePort: 9200