Issue #13 - SFC examples added

examples/README.md

@@ -0,0 +1,233 @@
+# Examples
+
+All presented below examples can be run on previously created Kubernetes Cluster (or single Node + Master).   
+To setup such cluster you should use this blueprint: https://github.com/cloudify-examples/simple-kubernetes-blueprint/tree/4.0.1
+
+
+### Simple example
+
+*simple-example-blueprint.yaml*
+
+TODO
+
+### Replicasets
+
+*replicasets-example-blueprint.yaml*
+
+TODO
+
+### Persistent volumes
+
+*persistent_volumes-example-blueprint*
+
+TODO
+
+### Service chaining
+
+There are 3 blueprints defined as examples of container-based service chaining for kubernetes.
+These scenarios are using Linux bridging and static routing to provide chain connectivity between separate pods.
+Implementations of all scenarios are done using utilities-plugin.
+Separate generic blueprint *vnf-blueprint* is used to define each pod and network interfaces by *service_chain* deployment.
+In other words main (*service_chain*) deployment is creating separate deployments from *vnf-blueprint* for each pod using utilities-plugin.
+
+So, before start you need to upload *vnf-blueprint* to Cloudify Manager. You can do it by:
+
+```
+cfy blueprints upload vnf-blueprint.yaml -b service_chain_vnf_component
+```
+
+It is also need to upload wagons for plugins used in blueprints:
+
+```
+cfy plugins upload https://github.com/cloudify-incubator/cloudify-utilities-plugin/releases/download/1.2.5/cloudify_utilities_plugin-1.2.5-py27-none-linux_x86_64-centos-Core.wgn
+cfy plugins upload https://github.com/cloudify-incubator/cloudify-kubernetes-plugin/releases/download/1.0.0/cloudify_kubernetes_plugin-1.0.0-py27-none-linux_x86_64.wgn
+cfy plugins upload http://repository.cloudifysource.org/cloudify/wagons/cloudify-fabric-plugin/1.5/cloudify_fabric_plugin-1.5-py27-none-linux_x86_64-centos-Core.wgn
+
+```
+
+Last step before making a deployments is to provide Kubernetes API credentials as a secrets.
+Using this approach the will be reusable for all deployments.
+You can do it executing:
+
+```
+cfy secrets create kubernetes_master_ip -s [IP ADDRESS OF KUBERNETES API]
+cfy secrets create kubernetes_master_user -s [SSH USERNAME FOR KUBERNETES MASTER]
+cfy secrets create kubernetes_master_ssh_key_path -s  [SSH KEY FILE PATH FOR KUBERNETES MASTER]
+```
+
+
+#### Example 1
+
+*service_chain_1-example-blueprint.yaml*
+
+Use case deploys chain with 3 containers:
+* client
+* VNF (router)
+* server
+
+![sfc_uc1](https://user-images.githubusercontent.com/20417307/28112813-b29b6a5c-66fa-11e7-8ecd-8c219a984412.jpg)
+
+You can deploy it executing:
+
+```
+cfy install -b service_chain_1 service_chain_1-example-blueprint.yaml
+```
+
+You can verify if this setup has been deployed correctly on Kuberentes VM using command line:
+
+1.  Check if all pod has been created. Execute:
+
+*kubectl get pods*
+
+You should see 3 pods. All have to be in 'Running' state:
+
+```
+NAME           READY     STATUS    RESTARTS   AGE
+client         1/1       Running   0          2m
+router         1/1       Running   0          2m
+server         1/1       Running   0          1m
+```
+
+2. Attach to 'client' console:
+
+*kubectl attach client -it*
+
+3. Perform tests for server connectivity ping ICMP traffic should pass
+
+*ping 192.168.1.7*
+
+4. Try to establish a ssh session. You should have possibility of making a connection.
+
+*ssh [email protected]*
+
+password: *test*
+
+5. Check if HTTP server is responding.
+
+*curl 192.168.1.7:8080*
+
+HTTP traffic should pass. A standard python SimpleHTTPServer directory listing should be present.
+
+*curl 192.168.1.7:8080/?q=banned*
+
+Expected 404 error.
+
+
+#### Example 2
+
+*service_chain_2-example-blueprint.yaml*
+
+Use case deploys chain with 4 containers:
+* client
+* VNF (router)
+* VNF (firewall)
+* server
+
+![sfc_uc2](https://user-images.githubusercontent.com/20417307/28112823-b7632502-66fa-11e7-9851-0bdc96017a4a.jpg)
+
+You can deploy it executing:
+
+```
+cfy install -b service_chain_2 service_chain_2-example-blueprint.yaml
+```
+
+You can verify if this setup has been deployed correctly on Kuberentes VM using command line:
+
+1.  Check if all pod has been created. Execute:
+
+*kubectl get pods*
+
+You should see 4 pods. All have to be in 'Running' state:
+
+```
+NAME           READY     STATUS    RESTARTS   AGE
+client         1/1       Running   0          2m
+router         1/1       Running   0          2m
+firewall       1/1       Running   0          2m
+server         1/1       Running   0          1m
+```
+
+2. Attach to 'client' console:
+
+*kubectl attach client -it*
+
+3. Perform tests for server connectivity ping ICMP traffic should pass
+
+*ping 192.168.1.7*
+
+4. Try to establish a ssh session. 
+TCP SSH traffic should be blocked by firewall.
+Making new connection should be impossible.
+
+*ssh [email protected]*
+
+5. Check if HTTP server is responding.
+
+*curl 192.168.1.7:8080*
+
+HTTP traffic should pass. A standard python SimpleHTTPServer directory listing should be present.
+
+*curl 192.168.1.7:8080/?q=banned*
+
+Expected 404 error.
+
+
+#### Example 3
+
+*service_chain_3-example-blueprint.yaml*
+
+Use case deploys chain with 5 containers:
+* client
+* VNF (router)
+* VNF (firewall)
+* VNF (URL filter)
+* server
+
+![sfc_uc3](https://user-images.githubusercontent.com/20417307/28112833-be9eb232-66fa-11e7-8ab5-dbdca51bda99.jpg)
+
+You can deploy it executing:
+
+```
+cfy install -b service_chain_3 service_chain_3-example-blueprint.yaml
+```
+
+You can verify if this setup has been deployed correctly on Kuberentes VM using command line:
+
+1.  Check if all pod has been created. Execute:
+
+*kubectl get pods*
+
+You should see 5 pods. All have to be in 'Running' state:
+
+```
+NAME           READY     STATUS    RESTARTS   AGE
+client         1/1       Running   0          2m
+router         1/1       Running   0          2m
+firewall       1/1       Running   0          2m
+filter         1/1       Running   0          2m
+server         1/1       Running   0          1m
+```
+
+2. Attach to 'client' console:
+
+*kubectl attach client -it*
+
+3. Perform tests for server connectivity ping ICMP traffic should pass
+
+*ping 192.168.1.7*
+
+4. Try to establish a ssh session. 
+TCP SSH traffic should be blocked by firewall.
+Making new connection should be impossible.
+
+*ssh [email protected]*
+
+5. Check if HTTP server is responding.
+
+*curl 192.168.1.7:8080*
+
+HTTP traffic should pass. A standard python SimpleHTTPServer directory listing should be present.
+
+*curl 192.168.1.7:8080/?q=banned*
+
+HTTP traffic should pass. A web page with information about a banned request is displayed

examples/common/docker.yaml

@@ -0,0 +1,17 @@
+node_types:
+  cloudify.docker.ImageBuilder:
+    derived_from: cloudify.nodes.Root
+    properties:
+      name:
+        type: string
+      dockerfile_content:
+        type: string
+    interfaces:
+      cloudify.interfaces.lifecycle:
+        create:
+          implementation: fabric.fabric_plugin.tasks.run_task
+          inputs:
+            tasks_file:
+              default: scripts/docker/image_builder.py
+            task_name:
+              default: create

examples/common/networking.yaml

@@ -0,0 +1,72 @@
+node_types:
+  cloudify.kubernetes.networking.linux.ConnectionPoint:
+    derived_from: cloudify.nodes.Root
+    properties:
+      pod_name:
+        type: string
+      name:
+        type: string
+      ip:
+        type: string
+      down:
+        type: boolean
+        default: false
+    interfaces:
+      cloudify.interfaces.lifecycle:
+        create:
+          implementation: fabric.fabric_plugin.tasks.run_task
+          inputs:
+            tasks_file:
+              default: scripts/networking/linux/connection_point.py
+            task_name:
+              default: create
+
+  cloudify.kubernetes.networking.linux.VirtualLink:
+    derived_from: cloudify.nodes.Root
+    properties:
+      name:
+        type: string
+      ip:
+        type: string
+      input_interface:
+        type: string
+      output_interface:
+        type: string
+    interfaces:
+      cloudify.interfaces.lifecycle:
+        create:
+          implementation: fabric.fabric_plugin.tasks.run_task
+          inputs:
+            tasks_file:
+              default: scripts/networking/linux/virtual_link.py
+            task_name:
+              default: create
+        delete:
+          implementation: fabric.fabric_plugin.tasks.run_task
+          inputs:
+            tasks_file:
+              default: scripts/networking/linux/virtual_link.py
+            task_name:
+              default: delete
+
+  cloudify.kubernetes.networking.linux.ForwardingPath:
+    derived_from: cloudify.nodes.Root
+    properties:
+      members:
+        description: ''
+    interfaces:
+      cloudify.interfaces.lifecycle:
+        create:
+          implementation: fabric.fabric_plugin.tasks.run_task
+          inputs:
+            tasks_file:
+              default: scripts/networking/linux/forwarding_path.py
+            task_name:
+              default: create
+        delete:
+          implementation: fabric.fabric_plugin.tasks.run_task
+          inputs:
+            tasks_file:
+              default: scripts/networking/linux/forwarding_path.py
+            task_name:
+              default: delete

examples/scripts/docker/image_builder.py

@@ -0,0 +1,56 @@
+from fabric.api import sudo
+from cloudify import ctx
+from cloudify.exceptions import NonRecoverableError
+
+
+def _get_input_parameter(name, kwargs):
+    parameter = ctx.node.properties.get(name, kwargs.get(name, None))
+
+    if not parameter:
+        raise NonRecoverableError(
+            'Mandatory input parameter {0} for lifecycle script not specified'
+            .format(name)
+        )
+
+    return parameter
+
+
+def _image_exisis(name):
+    ctx.logger.info('Checking if image {0} does exist'.format(name))
+
+    command = 'docker images {0} -q'.format(name)
+    ctx.logger.info('Executing command: {0}'.format(command))
+
+    return bool(sudo(command))
+
+
+def _build(name, dockerfile_content):
+    ctx.logger.info('Building image {0} ...'.format(name))
+
+    command = 'echo "" > {0}_temp.dockerfile'.format(name)
+    ctx.logger.info('Executing command: {0}'.format(command))
+    sudo(command)
+
+    for line in dockerfile_content:
+        command = 'echo "{0}" >> {1}_temp.dockerfile'.format(line, name)
+        ctx.logger.info('Executing command: {0}'.format(command))
+        sudo(command)
+
+    command = 'docker build -t {0} -f {0}_temp.dockerfile .'.format(name)
+    ctx.logger.info('Executing command: {0}'.format(command))
+    sudo(command)
+
+    ctx.logger.info('Build success')
+
+
+def create(**kwargs):
+    name = _get_input_parameter('name', kwargs)
+    dockerfile_content = _get_input_parameter('dockerfile_content', kwargs)
+
+    ctx.logger.info('Docker image builder started for image {0}'.format(name))
+
+    if not _image_exisis(name):
+        _build(name, dockerfile_content)
+        return
+
+    ctx.logger.warn('Image {0} already exists. Exiting.'.format(name))