odim

module

v0.0.0-...-85bdb93 Latest Latest Go to latest Published: Oct 8, 2021 License: Apache-2.0

Details

Valid go.mod file

The Go module system was introduced in Go 1.11 and is the official dependency management solution for Go.
Redistributable license

Redistributable licenses place minimal restrictions on how software can be used, modified, and redistributed.
Tagged version

Modules with tagged versions give importers more predictable builds.
Stable version

When a project reaches major version v1 it is considered stable.
Learn more about best practices

Repository

README ¶

Introduction

Resource Aggregator for Open Distributed Infrastructure Management (ODIM™) is a modular, open framework for simplified management and orchestration of distributed physical infrastructure.

Resource Aggregator for ODIM comprises the following two key components:

The resource aggregation function (the resource aggregator):

The resource aggregation function is the single point of contact between the northbound clients and the southbound infrastructure. Its primary function is to build and maintain a central resource inventory. It exposes Redfish-compliant APIs to allow northbound infrastructure management systems to:
- Get a unified view of the southbound compute, local storage, and Ethernet switch fabrics available in the resource inventory
- Gather crucial configuration information about southbound resources
- Manipulate groups of resources in a single action
- Listen to similar events from multiple southbound resources
One or more plugins:

The plugins abstract, translate, and expose southbound resource information to the resource aggregator through RESTful APIs. Resource Aggregator for ODIM supports:
- Generic Redfish plugin for ODIM (The GRF plugin): This plugin can be used for any Redfish-compliant device
- Dell plugin for ODIM: Plugin for managing Dell servers
- Plugin for unmanaged racks (URP): This plugin acts as a resource manager for unmanaged racks.
- Cisco ACI plugin: Plugin for managing Cisco ACI servers
- Integration of additional third-party plugins
Resource Aggregator for ODIM allows third parties to easily develop and integrate their plugins into its framework. For more information, see Resource Aggregator for Open Distributed Infrastructure Management™ Plugin Developer's Guide.

Filing Resource Aggregator for ODIM defects

Important: In case of any unforeseen issues you experience while deploying or using Resource Aggregator for ODIM, log on to the following website and file your defect by clicking Create.

Prerequisite: You must have valid LFN Jira credentials to create defects.

Website: https://jira.lfnetworking.org/secure/Dashboard.jspa
Discussion Forums: https://odim.slack.com/archives/C01DG9MH479
Documentation:
- Deployment Document- https://github.com/ODIM-Project/ODIM#readme
- Additional documents - https://github.com/ODIM-Project/ODIM/blob/main/docs

Resource Aggregator for ODIM deployment overview

Deploying Resource Aggregator for ODIM in a data center involves installing the following microservices on one or more machines:

Kubernetes microservices
The resource aggregator microservices:
- API
- Account-session
- Aggregation
- Events
- Fabrics
- Managers
- Systems
- Tasks
- Update
- Telemetry
The plugin microservices such as the Dell plugin, URP, and additional third-party plugins
Third-party services such as Kafka, etcd, Zookeeper, and Redis

These microservices can be deployed as portable, light-weight Docker containers. The containerized services are orchestrated and managed by Kubernetes—an open-source container orchestration platform that helps to automate, scale, and manage a containerized application. For more information on Kubernetes and its architecture, see https://kubernetes.io/docs/home/.

The following diagram illustrates how Resource Aggregator for ODIM is deployed and used in a Kubernetes environment. It indicates a cluster with three controller nodes (Node 1, Node 2 and Node 3) and that any additional worker nodes can be added into the cluster.

Deployment diagram

To deploy Resource Aggregator for ODIM, you will require:

One virtual machine (VM) or a physical machine called the deployment node to deploy Kubernetes and Resource Aggregator for ODIM microservices. You can deploy the Resource Aggregator for ODIM microservices using the odim-controller command-line utility and use commands to:
- Set up the Docker environment
- Set up a Kubernetes cluster
- Deploy the containerized Resource Aggregator for ODIM microservices and third-party services on the Kubernetes cluster nodes
- Manage the Resource Aggregator for ODIM deployment
One or more physical or virtual machines called cluster nodes where the containerized Resource Aggregator for ODIM microservices and third-party services are deployed as pods. The cluster nodes include controller and additional worker nodes to share the extra load. The controller node in a cluster also functions as a worker node. A cluster can have either one or three controller nodes. A cluster with three controller nodes provides a High Availability (HA) environment. In addition, you can add worker nodes into the cluster to scale up the resources and the services.

Each controller node has the following components:
- An underlying Ubuntu OS platform
- The Docker container engine
- The resource aggregator and the plugin microservice pods
- The infrastructure pods containing all the third-party services
- Kubelet, Kubeproxy, and the Kubernetes control plane comprising the API server, Scheduler, and the Controller-Manager
For more information on these Kubernetes components, see https://kubernetes.io/docs/concepts/overview/components/.

The following diagram is a logical representation of each controller node in a Kubernetes cluster.

Cluster node

The northbound management and orchestration systems access the Resource Aggregator for ODIM services through a virtual IP address (VIP) configured on the Kubernetes cluster using Keepalived. Communication between Resource Aggregator for ODIM and the southbound infrastructure happens through the same VIP.

Nginx acts as a reverse-proxy for the cluster nodes. Keepalived and Nginx together help in implementing high availability of the Resource Aggregator for ODIM services on the cluster nodes for both northbound management applications and southbound infrastructure to access.

Deployment considerations

The following is a list of considerations to be made while deploying Resource Aggregator for ODIM.

A deployment node is required to deploy Kubernetes and Resource Aggregator for ODIM microservices.
The following two deployment configurations are supported:
- One-node cluster: It has only one controller node that also functions as a worker node. It does not support scaling of the resources and services of Resource Aggregator for ODIM—you cannot add worker nodes into a one-node cluster.
- Three-node cluster: It has three controller nodes that also function as worker nodes for sharing the extra load. It provides HA environment by allowing the scaling of the Resource Aggregator for ODIM resources and services—you can add worker nodes and increase the number of service instances running in this cluster.
To convert an existing one-node cluster into a three-node cluster, you must reset the one-node deployment first and then modify the required parameters in the odim-controller configuration file.

NOTE: Resetting the existing deployment clears all data related to it.
Controller nodes of a Kubernetes cluster must not be removed.
The GRF plugin is not meant to be used in a production environment. Use it as reference while developing third-party plugins.
Scaling of the third-party services—Kafka and Redis clusters is not supported.
There must be at least one instance of a resource aggregator service and a plugin service running in the cluster. The maximum number of instances of a resource aggregator service and a plugin service that are allowed to run in a cluster is 10.

Resource Aggregator for ODIM compatibility matrix

The following table lists the software components and their versions that are compatible with Resource Aggregator for ODIM.

Software	Version
etcd	1.16
Java JRE	11
Kafka	2.5.0
Redis	6.2.5
Ubuntu LTS	18.04
ZooKeeper	3.5.7
Docker	19.03.8, build afacb8b7f0
Ansible	2.9.6
Kubespray	2.14.0
Helm charts	3.0.0
Nginx	1.14.0-0ubuntu1.7
Keepalived	1:1.3.9-1ubuntu0.18.04.2
Stakater/Reloader	v0.0.76
Redfish Schema	2021.1
Redfish Specification	1.11.1

Predeployment procedures

Set up the environment
Pull the Docker images of all the Kubernetes microservices
Build the Docker images of all the services
Generate an encrypted node password using the odim-vault tool
Configure log path for odim-controller

Setting up the environment

Hardware Prerequisites

Single deployment node with a minimum RAM of 8 GB (8192MB), three CPUs, and 100 GB of Hard Disk Drive (HDD)
Cluster nodes:
- To add 1,000 servers or less, you require nodes having 12 GB (12288 MB) RAM, 8 CPU cores and 16 threads, and 200 GB HDD each
- To add 5,000 servers or less, you require nodes having 32 GB (32768 MB) RAM, 16 CPU cores and 32 threads, and 200GB HDD each

Download and install Ubuntu 18.04 LTS on the deployment node and all the cluster nodes. During installation, configure the IP addresses of cluster nodes to reach the management VLANs where devices are connected. Ensure there is no firewall or switches blocking the connections and ports.

IMPORTANT: Ensure you create the same non-root username and password on all the cluster nodes during the installation of OS.
Verify that the time across all the nodes are synchronized. To know how to set up time sync, see Setting up time sync across nodes.

Run the following commands to install packages such as Python, Java, Ansible, and others on the deployment node.

NOTE: If the current version of a package is outdated, run the following commands to know the latest available version(s) of that package and install the first version listed in the output.

sudo apt-cache madison <package name>

Before running the following commands, ensure you are able to download the external packages through apt-get. If the nodes are behind a corporate proxy or firewall, set your proxy configuration on the deployment node and all the cluster nodes. To know how to set proxy, see Setting proxy configuration.

```
sudo apt-get update
```
```
sudo apt-get install sshpass=1.06-1 -y
```

sudo apt-get install python3.8=3.8.0-3~18.04.1 -y

sudo apt-get install python3-pip=9.0.1-2.3~ubuntu1.18.04.5 -y

sudo apt-get install software-properties-common=0.96.24.32.14 -y

sudo -E apt-add-repository ppa:ansible/ansible -y

sudo apt-get install openjdk-11-jre-headless=11.0.11+9-0ubuntu2~18.04 -y

```
python3 -m pip install --upgrade pip
```

sudo -H pip3 install ansible==2.9.6 --proxy=${http_proxy}

sudo -H pip3 install jinja2==2.11.1 --proxy=${http_proxy}

sudo -H pip3 install netaddr==0.7.19 --proxy=${http_proxy}

sudo -H pip3 install pbr==5.4.4 --proxy=${http_proxy}

sudo -H pip3 install hvac==0.10.0 --proxy=${http_proxy}

sudo -H pip3 install jmespath==0.9.5 --proxy=${http_proxy}

sudo -H pip3 install ruamel.yaml==0.16.10 --proxy=${http_proxy}

sudo -H pip3 install pyyaml==5.3.1 --proxy=${http_proxy}

Download and install go on the deployment node.
Configure Docker proxy on the deployment node.
Install Docker on the deployment node.
Perform the following steps to install Helm package on the deployment node:
1. Create a directory called helm to store the Helm tool installation script and navigate to it.
```
mkdir ~/helm
```
```
cd ~/helm
```
2. Fetch the latest Helm installation script.
```
curl -fsSL -o get_helm.sh https://raw.githubusercontent.com/helm/helm/master/scripts/get-helm-3 | /bin/bash
```
3. Change permissions of the Helm installation script file.
```
chmod 0700 get_helm.sh
```
4. Run the Helm installation script.
```
/bin/bash get_helm.sh
```

Pulling the Docker images of all the Kubernetes microservices

Run the following command on the deployment node to pull each Docker image in the following table:

docker pull <imagename>:<version>

Example: docker pull calico/cni:v3.15.1

The following table lists the details of all Docker images of the Kubernetes microservices to be pulled:

Docker image name	Version	Docker image file name
calico/cni	v3.15.1	calico_cni.tar
calico/kube-controllers	v3.15.1	calico_kube-controllers.tar
calico/node	v3.15.1	calico_node.tar
coredns/coredns	1.6.7	coredns_coredns.tar
k8s.gcr.io/cluster-proportional-autoscaler-amd64	1.8.1	k8s.gcr.io_cluster-proportional-autoscaler-amd64.tar
k8s.gcr.io/k8s-dns-node-cache	1.15.13	k8s.gcr.io_k8s-dns-node-cache.tar
k8s.gcr.io/kube-apiserver	v1.18.5	k8s.gcr.io_kube-apiserver.tar
k8s.gcr.io/kube-controller-manager	v1.18.5	k8s.gcr.io_kube-controller-manager.tar
k8s.gcr.io/kube-proxy	v1.18.5	k8s.gcr.io_kube-proxy.tar
k8s.gcr.io/kube-scheduler	v1.18.5	k8s.gcr.io_kube-scheduler.tar
k8s.gcr.io/pause	3.2	k8s.gcr.io_pause.tar
lachlanevenson/k8s-helm	v3.2.3	lachlanevenson_k8s-helm.tar
nginx	1.19	nginx.tar
quay.io/coreos/etcd	v3.4.3	quay.io_coreos_etcd.tar

Verify the images are successfully pulled using the following command.
```
docker images
```
You get an output similar to the following sample:
Save each Docker image to a tar archive using the following command:
```
docker save -o <Docker image file name> <Docker image name>
```
Example: docker save -o calico_node.tar calico/node
Copy each saved tar archive to a directory called kubernetes_images on the deployment node.

For example:cp /home/bruce/*.tar /home/bruce/kubernetes_images

IMPORTANT: When deploying ODIMRA, update the 'kubernetesImagePath' parameter in 'kube_deploy_nodes.yaml' file with the path of the 'kubernetes_images' directory you choose in this step. The images are automatically installed on all cluster nodes after deployment.

The 'kube_deploy_nodes.yaml' file is the configuration file used by odim-controller to set up a Kubernetes cluster and to deploy the Resource Aggregator for ODIM services.

NOTE: Verify the permissions of the archived tar files of the Docker images; the privilege of all files must be 'user:docker'.

Building Docker images of all the services

Run the following commands on the deployment node:

git clone https://github.com/ODIM-Project/ODIM.git

```
cd ODIM
```
```
export ODIMRA_USER_ID=2021
```
```
export ODIMRA_GROUP_ID=2021
```
```
./build_images.sh
```
```
sudo docker images
```
If the images are built successfully, you get an output similar to the following sample:
Pull the reloader and busybox images using the following commands:
```
docker pull stakater/reloader:v0.0.76

docker pull busybox:1.33
```

Save all Docker images in the following table to a tar archive using the following command:

docker save -o <image_name.tar> <image_name>:<version>

Example: docker save -o api.tar api:1.0

Docker image name	Version	Docker image bundle name
account-session	1.0	account-session.tar
aggregation	1.0	aggregation.tar
api	1.0	api.tar
events	1.0	events.tar
fabrics	1.0	fabrics.tar
managers	1.0	managers.tar
systems	1.0	systems.tar
task	1.0	task.tar
update	1.0	update.tar
kafka	1.0	kafka.tar
zookeeper	1.0	zookeeper.tar
etcd	1.16	etcd.tar
redis	1.0	redis.tar
stakater/reloader	v0.0.76	stakater_reloader.tar
busybox	1.33	busybox.tar
dellplugin	1.0	dellplugin.tar
urplugin	1.0	urplugin.tar
grfplugin	1.0	grfplugin.tar
telemetry	1.0	telemetry.tar

To install the Docker images of all services on the cluster nodes, create a directory called odimra_images on the deployment node and copy each tar archive to this directory. For example: cp /home/bruce/ODIM/*.tar /home/bruce/odimra_images

IMPORTANT: while deploying ODIMRA, update the 'odimraImagePath' parameter in 'kube_deploy_nodes.yaml' file with the path of the 'odimra_images' directory you choose in this step. The images are automatically installed on all cluster nodes after deployment.

NOTE: The 'kube_deploy_nodes.yaml' file is the configuration file used by odim-controller to set up a Kubernetes cluster and to deploy the Resource Aggregator for ODIM services.

Generating an encrypted node password

Encrypting the password of the local non-root user on the Kubernetes cluster nodes makes the deployment process non-interactive. If the encrypted password is not available during deployment, you are prompted to enter the password for the first time.

Resource Aggregator for ODIM uses the odim-vault tool to encrypt and decrypt passwords.

On the deployment node, navigate to ~/ODIM/odim-controller/scripts.
```
cd ~/ODIM/odim-controller/scripts
```

Build the odim-vault tool.

go build -ldflags "-s -w" -o odim-vault odim-vault.go

Enter a random string in a file called odimVaultKeyFile and save it.
```
vi odimVaultKeyFile
```
The entered string acts as the odim-vault crypto key. It is required for encrypting and decrypting the local user password of the Kubernetes cluster nodes.
To encode the entered odim-vault crypto key, run the following command:
```
./odim-vault -encode ~/ODIM/odim-controller/\
scripts/odimVaultKeyFile
```
Result: odimVaultKeyFile contains the encoded odim-vault master key.

Change the file permissions of odimVaultKeyFile.

chmod 0400 /home/${USER}/ODIM/odim-controller/\
scripts/odimVaultKeyFile

Enter the password of the default non-root user (that was set across all cluster nodes initially) in plain text in a file called nodePasswordFile. Save the file.
```
vi nodePasswordFile
```

To encrypt the entered password, run the following command:

./odim-vault -key ~/ODIM/odim-controller/\
scripts/odimVaultKeyFile -encrypt /home/${USER}/ODIM/odim-controller/\
scripts/nodePasswordFile

Result: nodePasswordFile contains the encrypted node password.

Change the file permissions of nodePasswordFile.

chmod 0400 /home/${USER}/ODIM/odim-controller/\
scripts/nodePasswordFile

Log path for odim-controller

You can provide a desired file path for the odim-controller logs by setting ODIM_CONTROLLER_LOG_PATH environment variable. Upon the execution of odim-controller, the log file will then be stored as {ODIM_CONTROLLER_LOG_PATH}/odim-controller.log.

If the environment variable is not set, the log file will be stored in the current directory where odim-controller was executed.

Upon each execution of odim-controller, the following information is logged:

Start time and end time of the execution
User and the groups, user belongs to
Options passed to odim-controller
Complete output of the operation

Deploying Resource Aggregator for ODIM and the Plugins

Topics covered in this section include:

Deploying the resource aggregator services
Deploying the Unmanaged Rack Plugin (URP)
Deploying the Dell plugin
Deploying the Cisco ACI plugin
Adding a plugin into the Resource Aggregator for ODIM framework

Deploying the resource aggregator services

Prerequisites

Ensure all the Predeployment procedures are complete.

Perform the following steps to update the odim-controller configuration file:

Navigate to ~/ODIM/odim-controller/scripts on the deployment node.
```
cd ~/ODIM/odim-controller/scripts
```
Copy content from the kube_deploy_nodes.yaml.tmpl file to the kube_deploy_nodes.yaml file by running the following command:
```
cp kube_deploy_nodes.yaml.tmpl kube_deploy_nodes.yaml
```

Edit the kube_deploy_nodes.yaml file.

vi kube_deploy_nodes.yaml

The kube_deploy_nodes.yaml file is the configuration file used by odim-controller to set up a Kubernetes cluster and to deploy the Resource Aggregator for ODIM services.

NOTE: Do not change the format of the content in this file. Doing so might interrupt the Resource Aggregator for ODIM deployment process.

When you open the kube_deploy_nodes.yaml file for the first time, it looks like the following (for a three node cluster):

#(C) Copyright [2020] Hewlett Packard Enterprise Development LP
# 
#Licensed under the Apache License, Version 2.0 (the "License"); you may
#not use this file except in compliance with the License. You may obtain
#a copy of the License at
# 
#    http://www.apache.org/licenses/LICENSE-2.0
# 
#Unless required by applicable law or agreed to in writing, software
#distributed under the License is distributed on an "AS IS" BASIS, WITHOUT
#WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. See the
#License for the specific language governing permissions and limitations
#under the License.

deploymentID:
httpProxy:
httpsProxy:
noProxy:
nodePasswordFilePath:
nodes:
  <Node1_Hostname>:
    ip: <Node1_IPAddress>
    username: <Node1_Username>
    priority: 100
  <Node2_Hostname>:
    ip: <Node2_IPAddress>
    username: <Node2_Username>
    priority: 99
  <Node3_Hostname>:
    ip: <Node3_IPAddress>
    username: <Node3_Username>
    priority: 98
odimControllerSrcPath:
odimVaultKeyFilePath:
odimCertsPath:
kubernetesImagePath:
odimraImagePath:
odimPluginPath:
odimra:
  groupID: 2021
  userID: 2021
  namespace: odim
  fqdn:
  rootServiceUUID:
  haDeploymentEnabled: True
  connectionMethodConf:
  - ConnectionMethodType: Redfish
    ConnectionMethodVariant: Compute:BasicAuth:GRF_v1.0.0
  etcHostsEntries:

  appsLogPath: /var/log/odimra
  odimraServerCertFQDNSan:
  odimraServerCertIPSan:
  odimraKafkaClientCertFQDNSan:
  odimraKafkaClientCertIPSan:

  apiProxyPort: 45000
  apiNodePort: 30080
  kafkaNodePort: 30092

  etcdDataPath: /etc/etcd/data
  etcdConfPath: /etc/etcd/conf

  kafkaConfPath: /etc/kafka/conf
  kafkaDataPath: /etc/kafka/data
  kafkaJKSPassword: "K@fk@_store1"

  redisOndiskDataPath: /etc/redis/data/ondisk
  redisInmemoryDataPath: /etc/redis/data/inmemory

  zookeeperConfPath: /etc/zookeeper/conf
  zookeeperDataPath: /etc/zookeeper/data
  zookeeperJKSPassword: "K@fk@_store1"

  nginxLogPath: /var/log/nginx
  virtualRouterID: 100
  virtualIP:

  rootCACert:
  odimraServerCert:
  odimraServerKey:
  odimraRSAPublicKey:
  odimraRSAPrivateKey:
  odimraKafkaClientCert:
  odimraKafkaClientKey:

For information on each parameter in this configuration file, see Odim-controller configuration parameters.

Update the following mandatory parameters in the above configuration file:

httpProxy (if your environment is behind a proxy)
httpsProxy (if your environment is behind a proxy)
noProxy (if your environment is behind a proxy)
deploymentID
nodePasswordFilePath
nodes (details of the single deployment node or the cluster nodes based on the type of your deployment)

For three node deployment:
- hostnames of node 1, node 2 and node 3
- IP address of node 1, node 2 and node 3
- username of node 1, node 2 and node 3
- Priority values of node 1, node 2 and node 3 (mandatory if haDeploymentEnabled is set to true)
odimControllerSrcPath
odimVaultKeyFilePath
odimraImagePath
odimPluginPath
fqdn
rootServiceUUID
connectionMethodConf
etcHostsEntries
apiProxyPort (mandatory if haDeploymentEnabled is set to true)
nginxLogPath (mandatory if haDeploymentEnabled is set to true)
virtualRouterID (mandatory if haDeploymentEnabled is set to true)
virtualIP (mandatory if haDeploymentEnabled is set to true)

Other parameters can either be empty or have default values. Optionally, you can update them with values based on your requirements.

NOTE: All parameters in the `kube_deploy_nodes.yaml` file get sorted alphabetically after the successful deployment of Resource Aggregator for ODIM services.

Sample of a "kube_deploy_nodes.yaml" file

#(C) Copyright [2020] Hewlett Packard Enterprise Development LP
# 
#Licensed under the Apache License, Version 2.0 (the "License"); you may
#not use this file except in compliance with the License. You may obtain
#a copy of the License at
# 
#    http://www.apache.org/licenses/LICENSE-2.0
# 
#Unless required by applicable law or agreed to in writing, software
#distributed under the License is distributed on an "AS IS" BASIS, WITHOUT
#WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. See the
#License for the specific language governing permissions and limitations
#under the License.

deploymentID: <Unique identifier for the deployment>
httpProxy: <HTTP Proxy to be set in the nodes>
httpsProxy: <HTTPS Proxy to be set in the nodes>
noProxy: <NO PROXY env to be set in the nodes>
nodePasswordFilePath: <Absolute path of the file containing encrypted node password>
nodes:
  knode1:
    ip: 10.18.24.100
    username: user
    priority: 100
  knode2:
    ip: 10.18.24.101
    username: user
    priority: 99
  knode3:
    ip: 10.18.24.102
    username: user
    priority: 98
odimControllerSrcPath: /home/user/ODIM/odim-controller
odimVaultKeyFilePath: /home/user/ODIM/odim-controller/scripts/odimVaultKeyFile
odimCertsPath: ""
kubernetesImagePath: /home/bruce/kubernetes_images
odimraImagePath: /home/bruce/odimra_images
odimPluginPath: ""
odimra:
  groupID: 2021
  userID: 2021
  namespace: odim
  fqdn: "odim.example.com"
 rootServiceUUID: "31991fcb-1ce9-4ca9-ab99-1d8181fcaa60"
  haDeploymentEnabled: True
  connectionMethodConf:
  - ConnectionMethodType: Redfish
    ConnectionMethodVariant: Compute:BasicAuth:GRF_v1.0.0
  - ConnectionMethodType: Redfish
    ConnectionMethodVariant: Storage:BasicAuth:STG_v1.0.0
  etcHostsEntries: ""

  appsLogPath: /var/log/odimra
  odimraServerCertFQDNSan: ""
  odimraServerCertIPSan: ""
  odimraKafkaClientCertFQDNSan: ""
  odimraKafkaClientCertIPSan: ""

  apiProxyPort: 45000
  apiNodePort: 30080
  kafkaNodePort: 30092

  etcdDataPath: /etc/etcd/data
  etcdConfPath: /etc/etcd/conf

  kafkaConfPath: /etc/kafka/conf
  kafkaDataPath: /etc/kafka/data
  kafkaJKSPassword: "K@fk@_store1"

  redisOndiskDataPath: /etc/redis/data/ondisk
  redisInmemoryDataPath: /etc/redis/data/inmemory

  zookeeperConfPath: /etc/zookeeper/conf
  zookeeperDataPath: /etc/zookeeper/data
  zookeeperJKSPassword: "K@fk@_store1"

  nginxLogPath: /var/log/nginx
  virtualRouterID: 100
  virtualIP: 10.18.24.103

  rootCACert:
  odimraServerCert:
  odimraServerKey:
  odimraRSAPublicKey:
  odimraRSAPrivateKey:
  odimraKafkaClientCert:
  odimraKafkaClientKey:

Set up a Kubernetes cluster by performing the following procedure:
1. Navigate to odim-controller/scripts on the deployment node:
```
cd ~/ODIM/odim-controller/scripts
```
2. Run the following command on the deployment node:
```
python3 odim-controller.py --deploy \
kubernetes --config /home/${USER}/ODIM/odim-controller/\
scripts/kube_deploy_nodes.yaml
```
3. Enable the non-root user to access the Kubernetes command-line tool (kubectl) on the cluster nodes. Run the following commands on each cluster node:
```
mkdir -p $HOME/.kube
```
```
sudo cp -i /etc/kubernetes/admin.conf $HOME/.kube/config
```
```
sudo chown $(id -u):$(id -g) $HOME/.kube/config
```
4. Verify that the Kubernetes pods are up and running in the cluster nodes. Run the following command on each cluster node:
```
kubectl get pods -n kube-system -o wide
```
  The following output is displayed.
Deploy the resource aggregator services:
1. Log in to the deployment node and run the following command:
```
python3 odim-controller.py --deploy \
odimra --config /home/${USER}/ODIM/odim-controller/\
scripts/kube_deploy_nodes.yaml
```
  All the resource aggregator services and the third-party services are successfully deployed.
2. Log in to each cluster node, run the following command on each cluster node to verify all deployed services are running successfully.
```
kubectl get pods -n odim -o wide
```
  Example output:
  
  If the services are not successfully deployed, reset the deployment and try deploying again. To reset, run the following command:
```
python3 odim-controller.py --reset odimra --config \
/home/${USER}/ODIM/odim-controller/scripts/kube_deploy_nodes.yaml \
--ignore-errors
```
  NOTE: - Resetting deployment removes the virtual IP configured through Keepalived. After reset, restart the Keepalived service.
  After resetting deployment, reset the odimCertsPath parameter to "" or to actual your certificate path in the kube_deploy_nodes.yaml file.
  
  IMPORTANT: Save the RootServiceUUID in the kube_deploy_nodes.yaml file in the path ~/ODIM/odim-controller/scripts/kube_deploy_nodes.yaml. If the services are not successfully deployed and you want to reset the deployment, you can use the saved RootServiceUUID.
Perform HTTP GET on /redfish/v1 using the following curl command on your deployment node.

IMPORTANT: Before running curl commands, unset http and https proxies, check if you have set proxy configuration. If yes, set "no_proxy" using the following commands:

unset https_proxy

unset http_proxy

export no_proxy="127.0.0.1,localhost,\
localhost.localdomain,10.96.0.0/12,\
<Comma-seperated_list_of_IP_addresses_of_the_deployment_node_and_the_cluster_nodes>"

Verify that all the resource aggregator services are listed in the JSON response body by running the following curl command:

curl -s --cacert \ 
{path_of_rootCA.crt} \ 
'https://{odim_host}:{port}/redfish/v1' -k

Replace {path_of_rootCA.crt} with the path specified for the odimCertsPath parameter in the kube_deploy_nodes.yaml file - <odimcertsPath>/rootCA.crt. The rootCA.crt file is required for secure SSL communication.
{odim_host} is the virtual IP address of the Kubernetes cluster.

NOTE: For a single node cluster configuration, {odim_host} is the ip address of master node. For a three node cluster configuration, to use FQDN as `{odim_host}`, ensure that FQDN is configured to the virtual IP address in the `/etc/hosts` file or in the DNS server.
{port} is the API server port configured in Nginx. The default port is 30080. If you have changed the default port, use that as the port.

The following JSON response is returned:

  {
   "@odata.context":"/redfish/v1/$metadata#ServiceRoot.ServiceRoot",
     	   "@odata.id":"/redfish/v1/",
     	   "@odata.type":"#ServiceRoot.v1_5_0.ServiceRoot",
     	   "Id":"RootService",
     	   "Registries":{
     	   "@odata.id":"/redfish/v1/Registries"
      	   },
            "SessionService":{
      	      "@odata.id":"/redfish/v1/SessionService"
      	   },
      	   "AccountService":{
      	      "@odata.id":"/redfish/v1/AccountService"
      	   },
      	   "EventService":{
      	      "@odata.id":"/redfish/v1/EventService"
      	   },
      	   "Tasks":{
      	      "@odata.id":"/redfish/v1/TaskService"
      	   },
      	   "TelemetryService":{
      	      "@odata.id":"/redfish/v1/TelemetryService"
      	   },
      	   "AggregationService":{
      	      "@odata.id":"/redfish/v1/AggregationService"
      	   },
      	   "Systems":{
      	      "@odata.id":"/redfish/v1/Systems"
      	   },
      	   "Chassis":{
      	      "@odata.id":"/redfish/v1/Chassis"
      	   },
      	   "Fabrics":{
      	      "@odata.id":"/redfish/v1/Fabrics"
      	   },
      	   "Managers":{
      	      "@odata.id":"/redfish/v1/Managers"
      	   },
      	   "UpdateService":{
      	      "@odata.id":"/redfish/v1/UpdateService"
      	   },
      	   "Links":{
      	      "Sessions":{
      	         "@odata.id":"/redfish/v1/SessionService/Sessions"
      	      }
      	   },
      	   "Name":"Root Service",
      	   "Oem":{
      	
      	   },
      	   "RedfishVersion":"1.11.1",
      	   "UUID":"0554d6ff-a7e7-4c94-80bd-da19125f95e5"
      	}

If you want to run curl commands on a different server, follow the instructions in Running curl commands on a different server.

Change the password of the default administrator account of Resource Aggregator for ODIM:

Username: admin

Password: Od!m12$4

To change the password, perform HTTP PATCH on the following URI:
```
https://{odim_host}:{port}/redfish/v1/AccountService/Accounts/{accountId}
```
Replace {accountId} with the username of the default administrator account.

Post the new password in a request body as shown in the sample request:

Sample request
```
{ 
   "Password":"Testing)9-_?{}"
}
```
Ensure that the new password meets the following requirements:
- Your password must not be same as your username.
- Your password must be at least 12 characters long and at most 16 characters long.
- Your password must contain at least one uppercase letter (A-Z), one lowercase letter (a-z), one digit (0-9), and one special character (~!@#$%^&*-+_|(){}:;<>,.?/). The default password is updated to the new password in the database.
To configure log rotation, perform the following procedure on each cluster node:
1. Navigate to the /etc/logrotate.d directory.
```
cd /etc/logrotate.d
```
2. Create a file called odimra.
3. Open the odimra file, add the following content and save:
```
/var/log/grfplugin_logs/*.log
/var/log/odimra/*.log
/opt/keepalived/logs/action_script.log
/opt/nginx/logs/error.log
/opt/nginx/logs/access.log {
    hourly
    missingok
    rotate 10
    notifempty
    maxsize 1M
    compress
    create 0644 odimra odimra
    shred
    copytruncate
}
```
  NOTE: After deploying a new plugin, log in to each cluster node and open the odimra file to add the log path entry for the new plugin.
4. Navigate to the /etc/cron.hourly directory.
```
cd /etc/cron.hourly
```
5. Create a file called logrotate.
6. Open the logrotate file and add the following content:
```
logrotate -s /var/lib/logrotate/status /etc/logrotate.d/odimra
```
7. To verify that the configuration is working, run the following command:
```
sudo logrotate -v -f /etc/logrotate.d/odimra
```

Deploying the Unmanaged Rack Plugin

Prerequisites

Kubernetes cluster is set up and the resource aggregator is successfully deployed.

Create a directory called plugins on the deployment node.
```
mkdir -p ~/plugins
```
In the plugins directory, create a directory called urplugin.
```
mkdir ~/plugins/urplugin
```

Log in to the deployment node and generate an encrypted password of Resource Aggregator for ODIM to be used in the urplugin-config.yaml file.

Run the following command and copy the output:

echo -n '<HPE ODIMRA password>' |openssl pkeyutl -encrypt -inkey <odimCertsPath>/odimra_rsa.private -pkeyopt rsa_padding_mode:oaep -pkeyopt rsa_oaep_md:sha512|openssl base64 -A

In this command, replace:

with the password of Resource Aggregator for ODIM (default administrator account password).
with the path you specified for the <odimCertsPath> parameter in the kube_deploy_nodes.yaml file.

Example output:

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

On the deployment node, copy the UR plugin configuration file and the hook script to ~/plugins/urplugin.

cp ~/ODIM/odim-controller/helmcharts/urplugin/urplugin-config.yaml ~/plugins/urplugin

cp ~/ODIM/odim-controller/helmcharts/urplugin/urplugin.sh ~/plugins/urplugin

Open the URP configuration YAML file.

vi ~/plugins/urplugin/urplugin-config.yaml

Sample urplugin-config.yaml file

urplugin:
  urPluginRootServiceUUID: e3473202-8706-4077-bd7d-d43d8d323a5b
  username: admin
  password: sTfTyTZFvNj5zU5Tt0TfyDYU-ye3_ZqTMnMIj-LAeXaa8vCnBqq8Ga7zV6ZdfqQCdSAzmaO5AJxccD99UHLVlQ==
  odimUsername: admin
  odimPassword: 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
  logPath: /var/log/urplugin_logs

Update the following parameters in the plugin configuration file:
- odimUsername: The username of the default administrator account of Resource Aggregator for ODIM.
- odimPassword: The encrypted password of the default administrator account of Resource Aggregator for ODIM. To generate the encrypted password, see step 3 of this procedure.
- urPluginRootServiceUUID: RootServiceUUID to be used by the URP service.
Other parameters can have default values. Optionally, you can update them with values based on your requirements. For more information on each parameter, see Plugin configuration parameters.
Generate Helm package for URP on the deployment node.
1. Navigate to odim-controller/helmcharts/urplugin.
```
cd ~/ODIM/odim-controller/helmcharts/urplugin
```
2. Run the following command to create urplugin Helm package at ~/plugins/urplugin:
```
helm package urplugin -d ~/plugins/urplugin
```
  The Helm package for URP is created in the tgz format.
Save the URP Docker image on the deployment node at ~/plugins/urplugin.
```
docker save urplugin:1.0 -o ~/plugins/urplugin/urplugin.tar
```
Navigate to the /ODIM/odim-controller/scripts directory on the deployment node.
```
cd ~/ODIM/odim-controller/scripts
```
Open the kube_deploy_nodes.yaml file by navigating to~/ODIM/odim-controller/scripts.
```
vi kube_deploy_nodes.yaml
```

Update the following parameters in the kube_deploy_nodes.yaml file to their corresponding values:

Parameter	Value
connectionMethodConf	The connection method associated with URP: ConnectionMethodVariant: `Compute:BasicAuth:URP_v1.0.0`
odimraKafkaClientCertFQDNSan	The FQDN to be included in the Kafka client certificate of Resource Aggregator for ODIM for deploying URP: `urplugin`, `api`. Add these values to the existing comma-separated list.
odimraServerCertFQDNSan	The FQDN to be included in the server certificate of Resource Aggregator for ODIM for deploying URP: `urplugin`, `api`. Add these values to the existing comma-separated list.
odimPluginPath	The path of the directory where the URP Helm package, the `urplugin` image, and the modified `urplugin-config.yaml` are copied.

Example:

       odimPluginPath: /home/bruce/plugins
        connectionMethodConf:
        - ConnectionMethodType: Redfish
          ConnectionMethodVariant: Compute:BasicAuth:GRF_v1.0.0
        - ConnectionMethodType: Redfish
          ConnectionMethodVariant: Compute:BasicAuth:URP_v1.0.0
           odimraKafkaClientCertFQDNSan: urplugin,api
           odimraServerCertFQDNSan: urplugin,api

Move odimra_kafka_client.key, odimra_kafka_client.crt, odimra_server.key and odimra_server.crt stored in odimCertsPath to a different folder.

NOTE: odimCertsPath is the absolute path of the directory where certificates required by the services of Resource Aggregator for ODIM are present. Refer to the "Odim-controller configuration parameters" section in this document for more information on odimCertsPath.

Upgrade odimra-secrets:

python3 odim-controller.py --config /home/${USER}/ODIM/odim-controller/scripts/kube_deploy_nodes.yaml --upgrade odimra-secret

Run the following command:

python3 odim-controller.py --config /home/${USER}/ODIM/odim-controller/scripts/kube_deploy_nodes.yaml --upgrade odimra-config

Run the following command to install Unmanaged Rack plugin:

python3 odim-controller.py --config /home/${USER}/ODIM/odim-controller/scripts/kube_deploy_nodes.yaml --add plugin --plugin urplugin

Run the following command on the cluster nodes to verify the Unmanaged Rack plugin pod is up and running:
```
kubectl get pods -n odim
```

Example output of the URP pod details:

NAME 				READY 	STATUS 		RESTARTS    AGE
urplugin-5fc4b6788-2xx97 	1/1 	Running 	0 	    4d22h

Add URP into the Resource Aggregator for ODIM framework.

Deploying the Dell plugin

Prerequisites

Kubernetes cluster is set up and the resource aggregator is successfully deployed.

Create a directory called plugins on the deployment node.
```
mkdir -p ~/plugins
```
In the plugins directory, create a directory called dellplugin.
```
mkdir ~/plugins/dellplugin
```

On the deployment node, copy the Dell plugin configuration file and the hook script to ~/plugins/dellplugin.

cp ~/ODIM/odim-controller/helmcharts/dellplugin/dellplugin-config.yaml ~/plugins/dellplugin

cp ~/ODIM/odim-controller/helmcharts/dellplugin/dellplugin.sh ~/plugins/dellplugin

Open the Dell plugin configuration YAML file.

vi ~/plugins/dellplugin/dellplugin-config.yaml

Sample dellplugin-config.yaml file:

dellplugin:
 eventListenerNodePort: 30084
 dellPluginRootServiceUUID: 7a38b735-8b9f-48a0-b3e7-e5a180567d37
 username: admin
 password: sTfTyTZFvNj5zU5Tt0TfyDYU-ye3_ZqTMnMIj-LAeXaa8vCnBqq8Ga7zV6ZdfqQCdSAzmaO5AJxccD99UHLVlQ==
 lbHost: 10.24.1.232
 lbPort: 30084
 logPath: /var/log/dellplugin_logs

Update the following parameters in the plugin configuration file:
- lbHost: IP address of the cluster node where the Dell plugin will be installed for one node cluster configuration. For three node cluster configuration, lbHost is the virtual IP address configured in Nginx and Keepalived.
- lbPort: Default port is 30084 for one node cluster configuration. For three node cluster configuration, lbport must be assigned with a free port (preferably above 45000) available on all cluster nodes. This port is used as nginx proxy port for the plugin.
  
  Note: The lbport is used as proxy port for eventlistenernodeport, which is used for subscribing to events.
- dellPluginRootServiceUUID: RootServiceUUID to be used by the Dell plugin service.
Other parameters can have default values. Optionally, you can update them with values based on your requirements. For more information on each parameter, see Plugin configuration parameters.
Generate the Helm package for the Dell plugin on the deployment node.
1. Navigate to odim-controller/helmcharts/dellplugin.
```
cd ~/ODIM/odim-controller/helmcharts/dellplugin
```
2. Run the following command to create dellplugin Helm package at ~/plugins/dellplugin:
```
helm package dellplugin -d ~/plugins/dellplugin
```
  The Helm package for the Dell plugin is created in the tgz format.
Save the Dell plugin Docker image on the deployment node at ~/plugins/dellplugin.
```
docker save dellplugin:1.0 -o ~/plugins/dellplugin/dellplugin.tar
```
Save the proxy configuration file install/templates/dellplugin_proxy_server.conf.j2 to ~/plugins/dellplugin.

Important: Do NOT change the value of any parameter in this file.
Navigate to the /ODIM/odim-controller/scripts directory on the deployment node.
```
cd ~/ODIM/odim-controller/scripts
```
Open the kube_deploy_nodes.yaml file.
```
 vi kube_deploy_nodes.yaml
```

Update the following parameters in the kube_deploy_nodes.yaml file to their corresponding values:

Parameter	Value
connectionMethodConf	The connection method associated with Dell plugin: ConnectionMethodVariant: `Compute:BasicAuth:DELL_v1.0.0`
odimraKafkaClientCertFQDNSan	The FQDN to be included in the Kafka client certificate of Resource Aggregator for ODIM for deploying the Dell plugin: `dellplugin`, `dellplugin-events` Add these values to the existing comma-separated list.
odimraServerCertFQDNSan	The FQDN to be included in the server certificate of Resource Aggregator for ODIM for deploying the Dell plugin: `dellplugin`, `dellplugin-events` Add these values to the existing comma-separated list.

Example:

odimPluginPath: /home/bruce/plugins
  connectionMethodConf:
  - ConnectionMethodType: Redfish
    ConnectionMethodVariant: Compute:BasicAuth:DELL_v1.0.0
  odimraKafkaClientCertFQDNSan: dellplugin,dellplugin-events
  odimraServerCertFQDNSan: dellplugin,dellplugin-events

Move odimra_kafka_client.key, odimra_kafka_client.crt, odimra_server.key and odimra_server.crt stored in odimCertsPath to a different folder.

NOTE: odimCertsPath is the absolute path of the directory where certificates required by the services of Resource Aggregator for ODIM are present. Refer to the "Odim-controller configuration parameters" section in this document for more information on odimCertsPath.

Upgrade odimra-secrets:

 python3 odim-controller.py --config /home/${USER}/ODIM/odim-controller/scripts/kube_deploy_nodes.yaml --upgrade odimra-secret

Run the following command:

 python3 odim-controller.py --config /home/${USER}/ODIM/odim-controller/scripts/kube_deploy_nodes.yaml --upgrade odimra-config

Run the following command to install the Dell plugin:

 python3 odim-controller.py --config /home/${USER}/ODIM/odim-controller/scripts/kube_deploy_nodes.yaml --add plugin --plugin dellplugin

Run the following command on the cluster nodes to verify the Dell plugin pod is up and running:
```
 kubectl get pods -n odim
```

Example output of the Dell plugin pod details:

NAME 				READY 	STATUS 		RESTARTS    AGE
dellplugin-5fc4b6788-2xx97 	1/1 	Running 	0 	    4d22h

Add the Dell plugin into the Resource Aggregator for ODIM framework.

Deploying the Cisco ACI plugin

Refer to the deployment instructions of the Cisco ACI plugin here.

Adding a plugin into the Resource Aggregator for ODIM framework

After a plugin is successfully deployed, you must add the plugin into the Resource Aggregator for ODIM framework to access the plugin service.

Prerequisites

The plugin you want to add is successfully deployed.

To add a plugin, perform HTTP POST on the following URI:

https://{odim_host}:{port}/redfish/v1/AggregationService/AggregationSources

{odim_host} is the virtual IP address of the Kubernetes cluster.
{port} is the API server port configured in Nginx. The default port is 30080. If you have changed the default port, use that as the port.

The following ports (except container ports) must be free:

Port name	Ports
Container ports (access restricted only to the Kubernetes cluster network)	45000 — API service port 45101- 45201 — Resource Aggregator for ODIM service ports 9082, 9092 — Kafka ports 6379 — Redis port 26379 — Redis Sentinel port 2181 — Zookeeper port 2379, 2380 — etcd ports
API node port (for external access)	30080
Kafka node port (for external access)	30092 for a one-node cluster configuration. 30092, 30093, and 30094 for a three-node cluster configuration
GRF plugin port EventListenerNodePort lbport	45001 — Port to be used while adding GRF plugin 30081 — Port used for event subscriptions in one-node cluster configuration lbport — For three-node cluster configuration, specify lbport as per your requirement. This port must be assigned with a free port (preferably above 45000) available on all cluster nodes. This port is used as nginx proxy port for the plugin For one-node cluster configuration, it is the same as EventListenerNodePort
UR plugin port	45007 — Port to be used while adding UR plugin
Dell plugin port EventListenerNodePort lbport	45005 — Port to be used while adding Dell plugin 30084 — Port used for event subscriptions in one-node cluster configuration lbport — For three-node cluster configuration, specify lbport as per your requirement. This port must be assigned with a free port (preferably above 45000) available on all cluster nodes. This port is used as nginx proxy port for the plugin. For one-node cluster configuration, it is the same as EventListenerNodePort

Provide a JSON request payload specifying:

The plugin address (the plugin name or hostname and the plugin port)
The username and password of the plugin user account
A link to the connection method having the details of the plugin

Sample request payload for adding the GRF plugin:

{
   "HostName":"grfplugin:45001",
   "UserName":"admin",
   "Password":"GRFPlug!n12$4",
   "Links":{
           "ConnectionMethod": {
             "@odata.id": "/redfish/v1/AggregationService/ConnectionMethods/d172e66c-b4a8-437c-981b-1c07ddfeacaa"
         }
   }
}

Sample request payload for adding URP:

{
   "HostName":"urplugin:45007",
   "UserName":"admin",
   "Password":"Plug!n12$4",
   "Links":{
           "ConnectionMethod": {
             "@odata.id": "/redfish/v1/AggregationService/ConnectionMethods/d172e66c-b4a8-437c-981b-1c07ddfeacaa"
         }
   }
}

Sample request payload for adding Dell:

{
   "HostName":"dellplugin:45005",
   "UserName":"admin",
   "Password":"Plug!n12$4",
  "Links":{
          "ConnectionMethod": {
            "@odata.id": "/redfish/v1/AggregationService/ConnectionMethods/d172e66c-b4a8-437c-981b-1c07ddfeacaa"
        }
  }
}

Request payload parameters

Parameter	Type	Description
HostName	String (required)	It is the plugin service name and the port specified in the Kubernetes environment. For default plugin ports, see Resource Aggregator for ODIM default ports.
UserName	String (required)	The plugin username. See default administrator account usernames of all the plugins in "Default plugin credentials".
Password	String (required)	The plugin password. See default administrator account passwords of all the plugins in "Default plugin credentials".
ConnectionMethod	Array (required)	Links to the connection methods that are used to communicate with this endpoint: `/redfish/v1/AggregationService/AggregationSources`. NOTE: Ensure that the connection method information for the plugin you want to add is updated in the odim-controller configuration file. To know which connection method to use, do the following: 1. Perform HTTP `GET` on: `/redfish/v1/AggregationService/ConnectionMethods`. You will receive a list of links to available connection methods. 2. Perform HTTP `GET` on each link. Check the value of the `ConnectionMethodVariant` property in the JSON response. It displays the details of a plugin. Choose a connection method having the details of the plugin of your choice. For available connection method variants, see the following "Connection method variants" table.

Plugin	Default username	Default password	Connection method variant
GRF plugin	admin	GRFPlug!n12$4	Compute:BasicAuth:GRF_v1.0.0
URP	admin	Plug!n12$4	Compute:BasicAuth:URP_v1.0.0

Use the following curl command to add the plugin:

curl -i POST \
   -H 'Authorization:Basic {base64_encoded_string_of_<odim_username:odim_password>}' \
   -H "Content-Type:application/json" \
   -d \
'{"HostName":"{plugin_host}:{port}",
  "UserName":"{plugin_userName}",
  "Password":"{plugin_password}", 
  "Links":{
      "ConnectionMethod": {
         "@odata.id": "/redfish/v1/AggregationService/ConnectionMethods/{ConnectionMethodId}"
      }
   }
}' \
 'https://{odim_host}:30080/redfish/v1/AggregationService/AggregationSources' -k

NOTE: To generate a base64 encoded string of `{odim_username:odim_password}`, run the following command:

echo -n '{odim_username}:{odim_password}' | base64 -w0

Replace {base64_encoded_string_of_[odim_username:odim_password]} with the generated base64 encoded string in the curl command. You will receive:

An HTTP 202 Accepted status code.
A link to the task monitor associated with this operation in the response header.

To know the status of this task, perform HTTP GET on the taskmon URI until the task is complete. If the plugin is added successfully, you will receive an HTTP 200 OK status code.

After the plugin is successfully added, it will also be available as a manager resource at:

/redfish/v1/Managers

For more information, refer to "Adding a plugin" in the Resource Aggregator for Open Distributed Infrastructure Management™ API Reference and User Guide.

To verify that the added plugin is active and running, do the following:
1. To get the list of all available managers, perform HTTP GET on:
  
  /redfish/v1/Managers
  
  You will receive JSON response having a collection of links to the manager resources. You will see the following links in the collection:
  - A link to the resource aggregator manager.
  - Links to all the added plugin managers.
2. To identify the plugin Id of the added plugin, perform HTTP GET on each manager link in the response.
  
  The JSON response body for a plugin manager has Name as the plugin name. Example: The JSON response body for the URP plugin manager has Name as URP.
  
  Sample response (URP manager)
```
{
   "@odata.context":"/redfish/v1/$metadata#Manager.Manager",
   "@odata.etag":"W/\"AA6D42B0\"",
   "@odata.id":"/redfish/v1/Managers/536cee48-84b2-43dd-b6e2-2459ac0eeac6",
   "@odata.type":"#Manager.v1_3_3.Manager",
   "FirmwareVersion":"1.0",
   "Id":"a9cf0e1e-c36d-4d5b-9a31-cc07b611c01b",
   "ManagerType":"Service",
   "Name":"URP",
   "Status":{
      "Health":"OK",
      "State":"Enabled"
   },
   "UUID":"a9cf0e1e-c36d-4d5b-9a31-cc07b611c01b"
}
```
Check in the JSON response of the plugin manager, if:
- State is Enabled
- Health is Ok
For more information, refer to "Managers" in Resource Aggregator for Open Distributed Infrastructure Management™ API Reference and User Guide.

Use cases for Resource Aggregator for ODIM

Adding a server into the resource inventory

To add a server, perform HTTP POST on the following URI with the request payload having details such as:

The BMC address (IP address or hostname)
The username and password of the BMC user account
A link to the connection method having the details of the plugin of your choice

URI: /redfish/v1/AggregationService/Actions/AggregationSources

Before adding a server, generate a certificate for it using the root CA certificate of Resource Aggregator for ODIM. To use your own root CA certificate to generate a certificate, you must first append it to the existing root CA certificate.

NOTE: To add a server using FQDN, add the server IP address and FQDN under the "etcHostsEntries" parameter in the "kube_deploy_nodes.yaml" file on the deployment node and run the following command:
python3 odim-controller.py --config \
/home/${USER}/ODIM/odim-controller/scripts/kube_deploy_nodes.yaml \
--upgrade configure-hosts

This action discovers information about a server and performs a detailed inventory of it. After successful completion, you will receive an aggregation source Id of the added BMC. Save it as it is required to identify it in the resource inventory later.

After the server is successfully added as an aggregation source, it will also be available as a computer system resource at /redfish/v1/Systems/ and a manager resource at /redfish/v1/Managers/.

For more information such as curl command, sample request, and sample response, see "Adding a server" in Resource Aggregator for Open Distributed Infrastructure Management™ API Reference and User Guide.

Viewing the resource inventory

To view the collection of servers available in the resource inventory, perform HTTP GET on the following URI:

/redfish/v1/Systems

For more information such as curl command, sample request, and sample response, see "Collection of computer systems" in Resource Aggregator for Open Distributed Infrastructure Management™ API Reference and User Guide.

Configuring BIOS settings for a server

To configure BIOS settings for a specific server, perform HTTP PATCH on the following URI with the request payload having BIOS attributes that you want to configure:

/redfish/v1/Systems/{ComputerSystemId}/Bios/Settings

For more information such as curl command, sample request, and sample response, see "Changing BIOS settings" in Resource Aggregator for Open Distributed Infrastructure Management™ API Reference and User Guide.

Resetting a server

To reset a specific server, perform HTTP POST on the following URI with the request payload specifying the type of reset such as ForceOn, ForceOff, On, ForceRestart, and more.

/redfish/v1/Systems/{ComputerSystemId}/Actions/ComputerSystem.Reset

To reset a group of servers, perform HTTP POST on the following URI with the request payload specifying the link and the type of reset for each server in the collection.

/redfish/v1/AggregationService/Actions/AggregationService.Reset

For more information such as curl command, sample request, and sample response, see "Resetting servers" and Resetting a computer system in Resource Aggregator for Open Distributed Infrastructure Management™ API Reference and User Guide.

Setting one time boot path for a server

To set boot path of a server, perform HTTP POST on the following URI:

/redfish/v1/Systems/{ComputerSystemId}/Actions/ComputerSystem.SetDefaultBootOrder

For more information such as curl command, sample request, and sample response, see "Changing the boot order of a computer system to default settings" in Resource Aggregator for Open Distributed Infrastructure Management™ API Reference and User Guide.

Searching the inventory for specific servers

To search servers in the inventory based on specific criteria, perform HTTP GET on the following URI. Specify the search criteria in the URI.

/redfish/v1/Systems?$filter={searchKeys}%20{conditionKeys}%20{value/regular_expression}%20{logicalOperand}%20{searchKeys}%20{conditionKeys}%20{value}

Example:

redfish/v1/Systems?filter=MemorySummary/TotalSystemMemoryGiB%20eq%20384

This URI searches the inventory for servers having total physical memory of 384 GB. On successful completion, it provides links to the filtered servers.

For more information such as curl command, sample request, and sample response, see "Search and filter" in Resource Aggregator for Open Distributed Infrastructure Management™ API Reference and User Guide.

Updating software and firmware

To upgrade or downgrade firmware of a system, perform HTTP POST on the following URIs:

/redfish/v1/UpdateService/Actions/UpdateService.SimpleUpdate
/redfish/v1/UpdateService/Actions/UpdateService.StartUpdate

Simple update action creates an update request or directly updates a software or a firmware component.

Start update action starts updating software or firmware components for which an update request has been created.

Subscribing to southbound events

To subscribe to events such as alerts and alarms from southbound resources and the resource aggregator, perform HTTP POST on the following URI with the request payload specifying the destination URI where events are received, the type of events such as Alert, ResourceRemoved, StatusChange, the links to the resources where events originate, and more.

/redfish/v1/EventService/Subscriptions

For more information such as curl command, sample request, and sample response, see "Creating an event subscription" in Resource Aggregator for Open Distributed Infrastructure Management™ API Reference and User Guide.

Viewing network fabrics

To view a collection of network fabrics and its switches, address pools, endpoints, and zones, perform HTTP GET on the following URIs respectively.

/redfish/v1/Fabrics

/redfish/v1/Fabrics/{fabricID}/Switches

/redfish/v1/Fabrics/{fabricID}/AddressPools

/redfish/v1/Fabrics/{fabricID}/Endpoints

/redfish/v1/Fabrics/{fabricID}/Zones

For more information such as curl command, sample request, and sample response, and for information on how to create fabric resources such as address pools, endpoints, and zones, see "Host to fabric networking" in Resource Aggregator for Open Distributed Infrastructure Management™ API Reference and User Guide.

Creating and deleting volumes

To create a volume, perform HTTP POST on the following URI with a request body specifying a name, the RAID type, and links to drives to contain the created volume:

/redfish/v1/Systems/{ComputerSystemId}/Storage/{storageSubsystemId}/Volumes

To remove an existing volume, perform HTTP DELETE on the following URI:

/redfish/v1/Systems/{ComputerSystemId}/Storage/{storageSubsystemId}/Volumes/{volumeId}

For more information such as curl command, sample request, and sample response, see "Creating a volume" and "Deleting a volume" in Resource Aggregator for Open Distributed Infrastructure Management™ API Reference and User Guide.

Removing a server from the resource inventory

To remove a server from the inventory, perform HTTP DELETE on the following URI with the request payload specifying a link to the server which you want to remove.

/redfish/v1/AggregationService/AggregationSources

This action erases the inventory of a specific server and also deletes all the event subscriptions associated with the server.

NOTE: You can remove only one server at a time.

For more information such as curl command, sample request, and sample response, see "Deleting a server" in [Resource Aggregator for Open Distributed Infrastructure Management™ API Reference and User Guide](https://github.com/ODIM-Project/ODIM/tree/development/docs).

Using odim-controller command-line interface

The odim-controller command-line interface (CLI) offers commands to support the following tasks:

Setting up the Kubernetes environment.
Deploying the services of Resource Aggregator for ODIM.
Other Kubernetes-related operations such as reset, upgrade, rollback, scale in and scale out, and more.

Command structure

python3 odim-controller.py [option(s)] [argument(s)]

Supported command options and arguments

Command option	Description
-h, --help	It provides information about a command.
--deploy	It is used to deploy a Kubernetes cluster or the services of Resource Aggregator for ODIM.Supported arguments: kubernetes, odimra
--reset	It is used to reset the existing Kubernetes deployment.Supported arguments: kubernetes, odimra
--addnode	It is used to add a node to an existing Kubernestes cluster.Supported arguments: kubernetes
--rmnode	It is used to remove a node from the existing Kubernetes cluster.Supported arguments: kubernetes
--config	It is used to specify the path of a configuration file.Supported arguments: Absolute path of a configuration file.
--dryrun	It is used to check configuration without deploying a Kubernetes cluster.
--noprompt	It is used to eliminate confirmation prompts.
--ignore-errors	It is used to ignore errors while resetting the Resource Aggregator for ODIM deployment.
--upgrade	It is used to upgrade the Resource Aggregator for ODIM deployment and configuration parameters. Supported arguments: odimra-config, odimra-platformconfig, configure-hosts, odimra-secret, kafka-secret, zookeeper-secret, account-session, aggregation, api, events, fabrics, managers, systems, task, update, kafka, zookeeper, redis, etcd, plugin, all, odimra, thirdparty NOTE:An upgrade operation takes a minimum of one minute to complete.
--scale	It is used to scale the deployment vertically—replicate the resource aggregator and plugin services.
--svc	Supported arguments: account-session, aggregation, api, events, fabrics, managers, systems, task, update, all
--plugin	It is used to provide the name of the plugin that you want to add, remove, scale, ot upgrade.Supported arguments: plugin
--add	It is used to add a plugin.Supported arguments: plugin
--remove	It is used to remove a plugin.Supported arguments: plugin
--replicas	It is used to specify the replica count of a service during scaling. Supported arguments: A number greater than zero and lesser than ten.
--list	Supported arguments: deployment, history
--rollback	It is used to roll back the deployment to a particular revision.
--revision	It is used to provide a revision number. It must be used with --rollback. NOTE: You can list revision numbers using the --list option.
--dep	It is used to provide the name of the deployment. It must be used in combination with the following options: - --list=history - --rollback

Example commands

 python3 odim-controller.py --addnode kubernetes --config \
 ~/ODIM/odim-controller/scripts/kube_deploy_nodes.yaml

  python3 odim-controller.py --config \
 ~/ODIM/odim-controller/scripts/kube_deploy_nodes.yaml \
 --scale --svc aggregation --replicas 3

For more examples, see Postdeployment operations.

Postdeployment operations

This section lists all the operations that you can perform after successfully deploying Resource Aggregator for ODIM. You can perform these operations to modify or upgrade the existing Kubernetes deployment.

NOTE: The operations listed in this section are not mandatory.

Scaling up the resources and services of Resource Aggregator for ODIM

Following are the two ways of scaling up the resources and services of Resource Aggregator for ODIM deployed in a Kubernetes cluster:

Horizontal scaling: It involves adding one or more worker nodes to the existing three-node cluster.

NOTE: Scaling of a one-node cluster is not supported—you cannot add nodes to a one-node cluster.
Vertical scaling: It involves creating multiple instances of the resource aggregator and plugin services.

NOTE:Scaling of third-party services is not supported.

To add a node, run the following command on the deployment node:
```
python3 odim-controller.py --addnode kubernetes --config \
/home/${USER}/ODIM/odim-controller/scripts/kube_deploy_nodes.yaml
```
Before adding a node, ensure that time on the node is same as the time on all the other existing nodes. To know how to set time sync, see Setting up time sync across nodes.
Log in to each cluster node and update all the configuration files inside /opt/nginx/servers with the new node details.
To scale up the resource aggregator deployments, run the following command on the deployment node:
```
python3 odim-controller.py --config \
/home/${USER}/ODIM/odim-controller/scripts/kube_deploy_nodes.yaml \
--scale --svc aggregation --replicas <replica_count>
```
Replace <deployment_name> with the name of the deployment which you want to scale up. To know all the supported deployment names, see Resource Aggregator for ODIM deployment names.

Replace <replica_count> with an integer indicating the number of service instances to be added.
To scale up the plugin services, run the following command on the deployment node:
```
python3 odim-controller.py --config \
/home/${USER}/ODIM/odim-controller/scripts/\
kube_deploy_nodes.yaml --scale --plugin <plugin_name> --replicas <replica_count>
```
Replace <plugin_name> with the name of the plugin whose service you want to scale up.

Replace <replica_count> with an integer indicating the number of plugin service instances to be added.

Scaling down the resources and services of Resource Aggregator for ODIM

Scaling down involves removing one or more worker nodes from an existing three-node cluster where the services of Resource Aggregator for ODIM are deployed.

NOTE: You cannot remove controller nodes in a cluster.

To remove a node, do the following:
1. Open the kube\_deploy\_nodes.yaml file on the deployment node.
2. Remove all the node entries under nodes except for the node that you want to remove.
3. Run the following command:
```
python3 odim-controller.py --rmnode kubernetes --config \
/home/${USER}/ODIM/odim-controller/scripts/kube_deploy_nodes.yaml
```
To scale down the resource aggregator deployments, run the following command on the deployment node:
```
python3 odim-controller.py --config \
/home/${USER}/ODIM/odim-controller/scripts/kube_deploy_nodes.yaml \
--scale --svc aggregation --replicas <replica_count>
```
Replace <deployment_name> with the name of the deployment which you want to scale down. To know all the supported deployment names, see Resource Aggregator for ODIM deployment names.

Replace <replica_count> with an integer indicating the number of service instances to be removed.
To scale down the plugin services, run the following command on the deployment node:
```
python3 odim-controller.py --config \
/home/${USER}/ODIM/odim-controller/scripts/kube_deploy_nodes.yaml \
--scale --plugin <plugin_name> --replicas <replica_count>
```
Replace <plugin_name> with the name of the plugin whose service you want to scale up.

Replace <replica_count> with an integer indicating the number of plugin service instances to be removed.

Rolling back to an earlier deployment revision

Rolling back the deployment of Resource Aggregator for ODIM to a particular revision restores the configuration manifest of that version.

To list the revision history of the deployment of Resource Aggregator for ODIM, run the following command:
```
python3 odim-controller.py --config \
/home/${USER}/ODIM/odim-controller/scripts/kube_deploy_nodes.yaml --list \
history --dep <deployment_name>
```
Replace <deployment_name> with the name of the deployment for which you want to list the revision history. To know all the supported deployment names, see HPE Resource Aggregator for ODIM deployment names.

You will receive a list of revisions along with the revision numbers.

To roll back the deployment of Resource Aggregator for ODIM to a particular revision, run the following command:

python3 odim-controller.py --config \
/home/${USER}/ODIM/odim-controller/scripts/kube_deploy_nodes.yaml --rollback \
--dep <deployment_name> --revision <revision_number>

Upgrading the Resource Aggregator for ODIM deployment

Upgrading the Resource Aggregator for ODIM deployment involves:

Updating the services of Resource Aggregator for ODIM to a new release.
Updating the configuration parameters of Resource Aggregator for ODIM.

NOTE: When you upgrade the Resource Aggregator for ODIM deployment, the new configuration manifests are saved by default.

To upgrade the resource aggregator deployments, run the following command:
```
python3 odim-controller.py --config /home/${USER}/ODIM/odim-controller/scripts/kube_deploy_nodes.yaml --upgrade <deployment_name>
```
Replace <deployment_name> with the name of the deployment which you want to upgrade. To know all the supported deployment names, see Resource Aggregator for ODIM deployment names.

To upgrade a plugin deployment, run the following command:

python3 odim-controller.py --config \
/home/${USER}/ODIM/odim-controller/scripts/kube_deploy_nodes.yaml
--upgrade plugin --plugin <plugin_name>

Replace <plugin_name> with the name of the plugin whose service you want to upgrade.

To update the odim-controller configuration parameters, do the following:
1. Navigate to ~/ODIM/odim-controller/scripts:
```
cd ~/ODIM/odim-controller/scripts
```
2. Open the kube_deploy_nodes.yaml file to edit:
```
vi kube_deploy_nodes.yaml
```
3. Edit the values of the parameters that you want to update and save the file.
  
  You cannot modify the following configuration parameters after the services of Resource Aggregator for ODIM are deployed.
  - appsLogPath
  - etcdConfPath
  - etcdDataPath
  - groupID
  - haDeploymentEnabled
  - hostIP
  - hostname
  - kafkaConfPath
  - kafkaDataPath
  - namespace
  - odimPluginPath
  - redisInmemoryDataPath
  - redisOndiskDataPath
  - rootServiceUUID
  - userID
  - zookeeperConfPath
  - zookeeperDataPath
4. Run the following command:
```
python3 odim-controller.py --config \
/home/${USER}/ODIM/odim-controller/scripts/kube_deploy_nodes.yaml \
--upgrade odimra-config
```

Appendix

Setting proxy configuration

Open the /etc/environment file to edit.
```
sudo vi /etc/environment
```

Add the following lines and save:

export http_proxy=<your_HTTP_proxy_address>
export https_proxy=<your_HTTP_proxy_address>
no_proxy="127.0.0.1,localhost,localhost.localdomain,10.96.0.0/12,<Deployment_Node_IP_address>,\
<Comma-separated-list-of-Ip-addresses-of-all-cluster-nodes>"

NOTE:When you add a node to the cluster, ensure to update no\_proxy with the IP address of the new node.

Example:

export http_proxy=<your_HTTP_proxy_address>
export https_proxy=<your_HTTP_proxy_address>
no_proxy="127.0.0.1,localhost,localhost.localdomain,10.96.0.0/12,<Deployment_Node_IP_address>,<Cluster_Node1_IP>,\
<Cluster_Node1_IP>,<Cluster_Node2_IP>,<Cluster_Node3_IP>"

Run the following command:
```
source /etc/environment
```

Setting up time sync across nodes

This procedure shows how to set up time synchronization across all the nodes of a Kubernetes cluster using Network Transfer Protocol (NTP).

Prerequisites

Ensure that all the nodes (deployment node and cluster nodes) are in the same time zone.

Open the timesyncd.conf file to edit:

sudo vi /etc/systemd/timesyncd.conf

Add the following lines and save:

[Time]

NTP=<NTP_server_IP_address>

#FallbackNTP=ntp.ubuntu.com

RootDistanceMaxSec=5

PollIntervalMinSec=1024

PollIntervalMaxSec=2048

Restart the systemd-timesyncd service using the following command:
```
sudo systemctl restart systemd-timesyncd
```
Check status to verify that the time sync is in place:
```
sudo systemctl status systemd-timesyncd
```

Downloading and installing Go

Run the following commands:

 wget https://dl.google.com/go/go1.13.7.linux-amd64.tar.gz -P /var/tmp

 sudo tar -C /usr/local -xzf /var/tmp/go1.13.7.linux-amd64.tar.gz

```
 export PATH=$PATH:/usr/local/go/bin
```

 mkdir -p ${HOME}/BRUCE/src ${HOME}/BRUCE/bin ${HOME}/BRUCE/pkg

```
 export GOPATH=${HOME}/BRUCE
```
```
 export GOBIN=$GOPATH/bin
```
```
 export GO111MODULE=on
```
```
 export GOROOT=/usr/local/go
```
```
 export PATH=$PATH:${GOROOT}/bin 
```

Configuring Docker proxy

NOTE: Before performing the following steps, ensure the `http_proxy`, `https_proxy`, and `no_proxy` environment variables are set.

[Optional] If the following content is not present in the /etc/environment file, add it:

cat << EOF | sudo tee -a /etc/environment
http_proxy=${http_proxy}
https_proxy=${https_proxy}
no_proxy=${no_proxy}
EOF

Run the following commands to update proxy information in the Docker service file:

sudo mkdir -p /etc/systemd/system/docker.service.d

cat << EOF | sudo tee /etc/systemd/system/docker.service.d/http-proxy.conf
[Service]
Environment="HTTP_PROXY=${http_proxy}"
Environment="HTTPS_PROXY=${https_proxy}"
Environment="NO_PROXY=${no_proxy}"
EOF

Run the following commands to update proxy information in the user Docker config file:

mkdir ~/.docker

sudo chown ${USER}:${USER} ~/.docker -R

sudo chmod 0700 ~/.docker

cat > ~/.docker/config.json <<EOF
 {
  "proxies":
  {
   "default":
   {
   "httpProxy": "${http_proxy}",
   "httpsProxy": "${https_proxy}",
   "noProxy": "${no_proxy}"
   }
 }
}
EOF

Installing Docker

Run the following commands:

NOTE: If the current version of a package is outdated, please run the following command to check for the latest available version of that package and install the first version listed in the output.

sudo apt-cache madison <package name>

sudo apt-get install -y apt-transport-https=1.6.12ubuntu0.2 ca-certificates=20210119~18.04.1 curl=7.58.0-2ubuntu3.14

sudo apt-get install -y gnupg-agent=2.2.4-1ubuntu1.3 software-properties-common=0.96.24.32.14

curl -fsSL https://download.docker.com/linux/ubuntu/gpg | sudo apt-key add -

sudo add-apt-repository "deb [arch=amd64] https://download.docker.com/linux/ubuntu $(lsb_release -cs) stable"

sudo apt-get install -y docker-ce=5:19.03.12~3-0~ubuntu-bionic docker-ce-cli=5:19.03.12~3-0~ubuntu-bionic containerd.io --allow-downgrades

Configure overlay storage for Docker:

cat << EOF | sudo tee /etc/docker/daemon.json
{
    "exec-opts": ["native.cgroupdriver=systemd"],
    "log-driver": "json-file",
    "storage-driver": "overlay2"
}
EOF

Perform the following steps to check and create Docker group if doesn't exist:
1. Check if the Docker group exists using the following command:
```
getent group docker
```
  Example output: docker:x:998:
2. Create the Docker group using the following command:
```
sudo groupadd docker
```
Configure to use Docker CLI without sudo access:
```
sudo usermod -aG docker $USER
```
Run the following command to activate the user added to the Docker group:
```
newgrp docker
```
NOTE: If you are unable to access Docker CLI without sudo even after performing this step, log out and log back in so that Docker group membership is re-evaluated.

Restart Docker service:

sudo systemctl enable docker

sudo systemctl restart docker

Verify that Docker is successfully installed:
```
docker run hello-world
```

Odim-controller configuration parameters

The following table lists all the configuration parameters required by odim-controller to deploy the services of Resource Aggregator for ODIM.

Parameter	Description
deploymentID	A unique identifier to identify the Kubernetes cluster. Example: "threenodecluster". It is required for the following operations: - Adding a node. - Deleting a node. - Resetting the Kubernetes cluster. - Deploying and removing the services of Resource Aggregator for ODIM.
httpProxy	HTTP Proxy to be set in all the nodes for connecting to external network. If there is no proxy available in your environment, you can leave it empty.
httpsProxy	HTTPS Proxy to be set in all the nodes for connecting to external network. If there is no proxy available in your environment, you can leave it empty.
noProxy	List of IP addresses and FQDNs for which proxy must not be used. It must begin with `127.0.0.1,localhost,localhost.localdomain,10.96.0.0/12,` followed by the IP addresses of the cluster nodes. If there is no proxy available in your environment, you can leave it empty.
nodePasswordFilePath	The absolute path of the file containing the encoded password of the nodes (encoded using the odim-vault tool) - `/home/<username\>/ODIM/odim-controller/scripts/nodePasswordFile`
nodes:	List of hostnames, IP addresses, and usernames of the nodes that are part of the Kubernetes cluster you want to set up. NOTE: For one-node cluster configuration, information of only the controller node is required.
Node_Hostname	Hostname of a cluster node. To know the hostname, run the following command on each node: `hostname`
ip	IP address of a cluster node.
username	Username of a cluster node. NOTE: Ensure that the username is same for all the nodes.
priority	An integer indicating the priority to be assigned to the Keepalived instance on a particular cluster node. A cluster node having the highest number as the priority value becomes the leader node of the cluster and the Virtual IP gets attached to it. For example, if there are three cluster nodes having the priority numbers as one, two, and three, the cluster node with the priority value of three becomes the leader node.
odimControllerSrcPath	The absolute path of the downloaded odim-controller source code - `/home/<username\>/ODIM/odim-controller`.
odimVaultKeyFilePath	The absolute path of the file containing the encrypted crypto key of the odim-vault tool - `/home/<username\>/ODIM/odim-controller/scripts/odimVaultKeyFile`
odimCertsPath	The absolute path of the directory where certificates required by the services of Resource Aggregator for ODIM are present. If you leave it empty, it gets updated to a default path during deployment (when odim-controller generates certificates required by the services of Resource Aggregator for ODIM). The default path of generated certificates is: `/home/<username>/ODIM/odim-controller/scripts/certs/<deploymentID\>` To generate and use your own CA certificates, see Using your own CA certificates and keys. Provide the path where you have stored your own CA certificates as the value for odimCertsPath.
kubernetesImagePath	Absolute path of the Kubernetes core images - `/home/<username>/ODIM/kubernetes_images`. NOTE: If it is left empty, the Kubernetes images will be downloaded from the Internet.
odimraImagePath	Absolute path of the images of Resource Aggregator for ODIM - `/home/<username>/ODIM/odimra_images`
odimPluginPath	Absolute path of the plugins directory - `/home/<username>/plugins`
odimra:	List of configurations required for deploying the services of Resource Aggregator for ODIM and third-party services.
groupID	Group ID to be used for creating the odimra group.The default value is 2021. You can optionally change it to a different value. NOTE: Ensure that the group id is not already in use on any of the nodes.
userID	User ID to be used for creating the odimra user. The default value is 2021. You can change it to a different value. NOTE: Ensure that the group id is not already in use on any of the nodes.
namespace	Namespace to be used for creating the service pods of Resource Aggregator for ODIM. The default value is "odim". You can optionally change it to a different value.
fqdn	Name of the server associated with the services of Resource Aggregator for ODIM. This name is used for communication among the services of Resource Aggregator for ODIM. Example: "odim.example.com".
rootServiceUUID	RootServiceUUID to be used by the resource aggregator and the plugin services. To generate an UUID, run the following command: `uuidgen` Copy the output and paste it as the value for rootServiceUUID.
haDeploymentEnabled	When set to true, it deploys third-party services as a three-instance cluster. By default, it is set to true.
connectionMethodConf	Parameters of type array required to configure the supported connection methods. NOTE: To deploy a plugin after deploying the resource aggregator services, add its connection method information in the array and update the file using odim-controller `--upgrade` option.
kafkaNodePort	The port to be used for accessing the Kafka services from external services. The default port is 30092. You can optionally change it. NOTE: Ensure that the port is in the range of 30000 to 32767.
etcHostsEntries	List of FQDNs of the external servers and plugins to be added to the `/etc/hosts` file in each of the service containers of Resource Aggregator for ODIM. The external servers are the servers that you want to add into the resource inventory. NOTE: It must be in the YAML multiline format as shown in the "etcHostsEntries template".
appsLogPath	The path where the logs of the Resource Aggregator for ODIM services must be stored. The default path is `/var/log/odimra`.
odimraServerCertFQDNSan	List of FQDNs to be included in the server certificate of Resource Aggregator for ODIM. It is required for deploying plugins. NOTE: When you add a plugin, add the FQDN of the new plugin to the existing comma-separated list of FQDNs.
odimraServerCertIPSan	List of IP addresses to be included in the server certificate of Resource Aggregator for ODIM. It is required for deploying plugins. NOTE: It must be comma-separated values of type String.
odimraKafkaClientCertFQDNSan	List of FQDNs to be included in the Kafka client certificate of Resource Aggregator for ODIM. It is required for deploying plugins. NOTE: When you add a plugin, add the FQDN of the new plugin to the existing comma-separated list of FQDNs.
odimraKafkaClientCertIPSan	List of IP addresses to be included in the Kafka client certificate of Resource Aggregator for ODIM. It is required for deploying plugins.
apiProxyPort	Any free port on the cluster node having high priority. It must be available on all the other cluster nodes. Preferred port is above 45000. Ensure that this port is not used as any other service port. NOTE: You can reach the resource aggregator API server at: `https://<VIP>:<nginx_api_port>`.
apiNodePort	The port to be used for accessing the API service of Resource Aggregator for ODIM.The default port is 30080. You can optionally use a different port. NOTE: Ensure that the port is in the range of 30000 to 32767.
etcdDataPath	The path to persist etcd data.
etcdConfPath	The path to store etcd configuration data.
kafkaConfPath	The path to store Kafka configuration data.
kafkaDataPath	The path to persist Kafka data.
kafkaJKSPassword	The password of the Kafka keystore.
redisOndiskDataPath	The path to persist on disk Redis data.
redisInmemoryDataPath	The path to persist in-memory Redis data.
zookeeperConfPath	The path to store Zookeeper configuration data.
zookeeperDataPath	The path to persist Zookeeper data.
zookeeperJKSPassword	The password of the ZooKeeper keystore.
nginxLogPath	The path where Nginx logs are stored.
virtualRouterID	A unique number acting as the virtual router ID. It must be in the range of 0 to 250. Ensure that it is same on all the cluster nodes, but different across deployments (if there is more than one deployment).
virtualIP	Any free Virtual IP address to be attached to the leader node of a cluster. It acts as the IP address of the cluster. Ensure that the chosen virtual IP address is not associated with any cluster node. The northbound client applications reach the resource aggregator API service through this virtual IP address.
rootCACert	The path of the Resource Aggregator for ODIM root CA certificate. It gets updated automatically during deployment.
odimraKafkaClientCert	The path of the Kafka client certificate. It gets updated automatically during deployment.
odimraKafkaClientKey	The path of the Kafka client key. It gets updated automatically during deployment.
odimraRSAPrivateKey	The path of the RSA private key. It gets updated automatically during deployment.
odimraRSAPublicKey	The path of the RSA public key. It gets updated automatically during deployment.
odimraServerKey	The path of the Resource Aggregator for ODIM server key. It gets updated automatically during deployment.

Note: The paramters priority, apiProxyPort, ngnixLogPath, virtualRouterID and virtualIP are mandatory only when haDeploymentEnabled is set to true.

etcHostsEntries template

etcHostsEntries: |
 <IP_address_of_external_server_or_plugin> <FQDN_of_external_server>

Example:

odimra:
 etcHostsEntries: |
  1.1.1.1 odim1.example.com
  2.2.2.2 odim2.example.com

Running curl commands on a different server

To run curl commands on a different server, perform the following steps to provide the rootCA.crt file.

Navigate to the path specified for the odimCertsPath parameter in the kube_deploy_nodes.yaml file on the deployment node.
Copy the rootCA.crt file.
Log in to your server and paste the rootCA.crt file in a folder.
Open the /etc/hosts file to edit.
Scroll to the end of the file, add the following line, and then save:
```
<hostvm_ipv4_address> <FQDN>
```

Check if curl is working by using the following command:

curl --cacert rootCA.crt 'https://{odim_host}:{port}/redfish/v1'

NOTE: - To avoid using the `--cacert` flag in every curl command, add `rootCA.crt` in the `ca-certificates.crt` file available in this path: `/etc/ssl/certs/ca-certificates.crt`. You can access the base URL using a REST client. To access it using a REST client, add the rootCA.crt file of HPE Resource Aggregator for ODIM to the browser where the REST client is launched.

## Plugin configuration parameters

The following table lists all the configuration parameters required to deploy a plugin service:

Parameter	Description
odimra	List of configurations required for deploying the services of Resource Aggregator for ODIM and third-party services. NOTE: Ensure that the values of the parameters listed under odimra are same as the ones specified in the `kube_deploy_nodes.yaml` file.
namespace	Namespace to be used for creating the service pods of Resource Aggregator for ODIM. The default value is "odim". You can optionally change it to a different value.
groupID	Group ID to be used for creating the odimra group. The default value is 2021. You can optionally change it to a different value. NOTE: Ensure that the group id is not already in use on any of the nodes.
haDeploymentEnabled	When set to true, it deploys third-party services as a three-instance cluster. By default, it is set to true. Before setting it to false, ensure that there are at least three nodes in the Kubernetes cluster.
eventListenerNodePort	The port used for listening to plugin events. Refer to the sample plugin yaml configuration file to view the sample port information.
RootServiceUUID	RootServiceUUID to be used by the plugin service. To generate a UUID, run the following command: `uuidgen` Copy the output and paste it as the value of the plugin's rootServiceUUID.
username	Username of the plugin.
password	The encrypted password of the plugin.
lbHost	If there is only one cluster node, the lbHost is the IP address of the cluster node. If there is more than one cluster node (haDeploymentEnabled is true), lbHost is the virtual IP address configured in Nginx and Keepalived.
lbPort	If it is a one-cluster configuration, the lbPort must be same as eventListenerNodePort. If there is more than one cluster node (haDeploymentEnabled is true), lbport must be assigned with a free port (preferably above 45000) available on all cluster nodes. This port is used as nginx proxy port for the plugin.
logPath	The path where the plugin logs are stored. The default path is `/var/log/<plugin_name>_logs` Example: `/var/log/grfplugin\_logs`

Resource Aggregator for ODIM deployment names

Deployment name	Description
odimra-config	Deployment name of the ConfigMap which contains the configuration information required by the resource aggregator services
odimra-platformconfig	Deployment name of the ConfigMap which contains the Kafka client configuration information required by the resource aggregator services
configure-hosts	Deployment name of the ConfigMap which contains the entries to be added in the `/etc/hosts` file on all the containers
odimra-secret	Deployment name of the secret which contains the certificates and keys used by the resource aggregator services
kafka-secret	Deployment name of the secret which contains the JKS password of Kafka keystore
zookeeper-secret	Deployment name of the secret which contains the JKS password of zookeeper keystore
account-session	Deployment name of the account-sessions service
aggregation	Deployment name of the aggregation service
api	Deployment name of the api service
events	Deployment name of the events service
fabrics	Deployment name of the fabrics service
managers	Deployment name of the managers service
systems	Deployment name of the systems service
tasks	Deployment name of the tasks service
update	Deployment name of the update service
kafka	Deployment name of the Kafka service
zookeeper	Deployment name of the zookeeper service
redis	Deployment name of the Redis service
etcd	Deployment name of the etcd service
all	Deployment name to be used for scaling up all the resource aggregator services
odimra	Deployment name to be used for scaling up all ConfigMaps, secrets, and the resource aggregator services
thirdparty	Deployment name to be used for scaling up all the third-party ConfigMaps and secrets

Using your own CA certificates and keys

Generate the following certificates and store them in a folder on the deployment node:
- odimra_rsa.private: It must be generated only once before deploying the resource aggregator and plugin services.
- odimra_rsa.public: It must be generated only once before deploying the resource aggregator and plugin services.
  
  NOTE: Ensure not to replace the RSA public and private keys. Replacing them results in loss of data and requires reinstallation. If you are generating your own CA certificates to replace the existing CA certificates, move the existing odimra\_rsa.public and odimra\_rsa.private files to the folder where you are generating all the other certificates.
- rootCA.crt: Root CA certificate.
- odimra_server.crt: Certificate to be used by the API gateway and the resource aggregator and plugin components.
- odimra_server.key: Private key to be used by the API gateway and the resource aggregator and plugin components.
- odimra_kafka_client.crt: Kafka certificate.
- odimra_kafka_client.key: Kafka key.
- kafka.keystore.jks: Keystore of type jks used by Kafka servers for TLS-based communication.
- kafka.truststore.jks: Truststore of type jks. It contains CA certificates, used by Kafka server to validate certificates of the client that is contacting the Kafka server.
- zookeeper.keystore.jks: Keystore of type jks used by Zookeeper for TLS-based communication.
- zookeeper.truststore.jks: Truststore of type jks. It contains CA certificates, used by Zookeeper to validate certificates of the client that is contacting it.
While generating these certificates:
- Ensure that odimra_server.crt has the SAN entry of FQDN specified in the kube_deploy_nodes.yaml file.
- Ensure that odim_server.crt and odim_kafka_client.crt have the following SAN entries for all the plugins you want to deploy.
  
  Examples: grfplugin: grfplugin,grfplugin-events.
- Ensure that kafka.truststore.jks and kafka.truststore.jks have the following SAN entries:
```
DNS.1 = kafka
DNS.2 = kafka1.kafka.${ODIMRA_NAMESPACE}.svc.cluster.local
DNS.3 = kafka2.kafka.${ODIMRA_NAMESPACE}.svc.cluster.local
DNS.4 = kafka3.kafka.${ODIMRA_NAMESPACE}.svc.cluster.local
DNS.5 = kafka-ext
DNS.6 = kafka1-ext
DNS.7 = kafka2-ext
DNS.8 = kafka3-ext
```
- Ensure that zookeeper.truststore.jks and zookeeper.truststore.jks have the following SAN entries:
```
DNS.1 = zookeeper
DNS.2 = zookeeper1.zookeeper.${ODIMRA_NAMESPACE}.svc.cluster.local
DNS.3 = zookeeper2.zookeeper.${ODIMRA_NAMESPACE}.svc.cluster.local
DNS.4 = zookeeper3.zookeeper.${ODIMRA_NAMESPACE}.svc.cluster.local
```
Replace {ODIMRA_NAMESPACE} with the value specified for namespace in the kube_deploy_nodes.yaml file.
Update odimCertsPath with the path of the folder where you have stored the certificates in the kube_deploy_nodes.yaml file.
[Optional] Perform this step only after you have successfully deployed the resource aggregator and plugin services.

If you want to replace the existing CA certificates, run the following command:

python3 odim-controller.py --config \
/home/${USER}/ODIM/odim-controller/scripts/kube_deploy_nodes.yaml --upgrade odimra-secret

Result: The existing certificates are replaced with the new certificates and the resource aggregator pods are restarted.

Regenerating certificates

Updating Kafka password and certificate

Open the ~/ODIM/odim-controller/scripts/kube_deploy_nodes.yaml file on the deployment node:
```
vi ~/ODIM/odim-controller/scripts/kube_deploy_nodes.yaml
```
Update the kafkaJKSPassword property with the new password and save.

NOTE:You might want to store this password for later—if there is a rollback, use the stored password.
Kafka JKS password is updated and Kafka certificate and key are regenerated.
Regenerate Kafka JKS along with the certificate and key:

Move the following files from the path mentioned in odimCertsPath property in the kube_deploy_nodes.yaml file to a different folder:
- kafka.keystore.jks
- kafka.truststore.jks

Update Kafka secret:

python3 odim-controller.py --config \
/home/${USER}/ODIM/odim-controller/scripts/kube_deploy_nodes.yaml --upgrade kafka-secret

Kafka secret is updated and Kafka pods are restarted.

Updating Zookeeper password and certificate

Open the ~/ODIM/odim-controller/scripts/kube_deploy_nodes.yaml file on the deployment node:
```
vi ~/ODIM/odim-controller/scripts/kube_deploy_nodes.yaml
```
Update the zookeeperJKSPassword property with the new password and save.

NOTE:You might want to store this password for later—if there is a rollback, use the stored password.
Zookeeper JKS password is updated and Zookeeper certificate and key are regenerated.
Regenerate Zookeeper JKS along with the certificate and key:

Move the following files from the path mentioned in odimCertsPath property in the kube_deploy_nodes.yaml file to a different folder:
- zookeeper.keystore.jks
- zookeeper.truststore.jks

Update Zookeeper secret:

python3 odim-controller.py --config \
/home/${USER}/ODIM/odim-controller/scripts/kube_deploy_nodes.yaml --upgrade zookeeper-secret

Zookeeper secret is updated and Zookeeper pods are restarted.

Updating certificates with SAN entries

To update the odimra-server.crt and odimra_kafka_client.crt files, do the following:

NOTE: You cannot update `odimra_rsa.private` and `odimra_rsa.public.

To add new entries in odimra-server.crt:
1. Open the ~/ODIM/odim-controller/scripts/kube_deploy_nodes.yaml file on the deployment node:
```
vi ~/ODIM/odim-controller/scripts/kube_deploy_nodes.yaml
```
2. Add the new IP and FQDN SANs to the odimraServerCertIPSan and odimraServerCertFQDNSan parameters respectively and save.
```
odimraServerCertIPSan: 1.1.1.1,<new_IP>
odimraServerCertFQDNSan: odim1.com,<new_FQDN>
```
3. Move odimra_server.key and odimra_server.crt stored in the path specified for odimCertsPath to a different folder.
To add new entries in odimra_kafka_client.crt:
1. Add the new IP and FQDN SANs to the odimraKafkaClientCertIPSan and odimraKafkaClientCertFQDNSan parameters respectively and save.
```
odimraKafkaClientCertIPSan: 1.1.1.1,<new_IP>
odimraKafkaClientCertFQDNSan: odim1.com,<new_FQDN>
```
2. Move odimra_kafka_client.key and odimra_kafka_client.crt stored in the path specified for odimCertsPath to a different folder.

Update odimra-secrets:

python3 odim-controller.py --config \
/home/${USER}/ODIM/odim-controller/scripts/kube_deploy_nodes.yaml --upgrade odimra-secret

All the Resource Aggregator for ODIM pods are restarted.

Updating `/etc/hosts` in the containers

Open the ~/ODIM/odim-controller/scripts/kube_deploy_nodes.yaml file on the deployment node:
```
vi ~/ODIM/odim-controller/scripts/kube_deploy_nodes.yaml
```
Update the etcHostsEntries with a new entry and save.

Run the following command:

python3 odim-controller.py --config \
/home/${USER}/ODIM/odim-controller/scripts/kube_deploy_nodes.yaml \
--upgrade configure-hosts

The /etc/hosts file is updated in all the containers.

Appending CA certificates to the existing Root CA certificate

Open the ~/ODIM/odim-controller/scripts/kube_deploy_nodes.yaml file on the deployment node:
```
vi ~/ODIM/odim-controller/scripts/kube_deploy_nodes.yaml
```
Copy the path specified for the odimCertsPath parameter in the kube_deploy_nodes.yaml file.
Append the CA certificate to the rootCA.crt file available in the odimCertsPath path:
```
cat CA.crt >> <odimCertsPath>/rootCA.crt
```

Update odimra-secrets:

python3 odim-controller.py --config \
/home/${USER}/ODIM/odim-controller/scripts/kube_deploy_nodes.yaml --upgrade odimra-secret

All the pods of Resource Aggregator for ODIM are restarted.

Resource Aggregator for ODIM default ports

Default ports used by the resource aggregator, plugins, and third-party services can be found in the table in the Adding a plugin into the Resource Aggregator for ODIM framework section.

Deploying the GRF plugin

Prerequisites

Kubernetes cluster is set up and the resource aggregator is successfully deployed.

Create a directory called plugins on the deployment node.
```
mkdir -p ~/plugins
```
In the plugins directory, create a directory called grfplugin.
```
mkdir ~/plugins/grfplugin
```

On the deployment node, copy the GRF configuration file and the hook script to ~/plugins/grfplugin.

cp ~/ODIM/odim-controller/helmcharts/grfplugin/grfplugin-config.yaml ~/plugins/grfplugin

cp ~/ODIM/odim-controller/helmcharts/grfplugin/grfplugin.sh ~/plugins/grfplugin

Open the GRF plugin configuration YAML file.

vi ~/plugins/grfplugin/grfplugin-config.yaml

Sample grfplugin-config.yaml file:

grfplugin:
  eventListenerNodePort: 30081
  rootServiceUUID: 65963042-6b99-4206-8532-dcd085a835b1
  username: admin
  password: "UUFCYFpBoHh6UdvytPzm65SkHj5zyl73EYVNJNbrFeAPWYrkpTijGB9zrVQSbbLv052HK7-7chqDQQcjgWf7YA=="
  lbHost: <Ngnix_virtual_IP_address>
  lbPort: <Ngnix_plugin_port>
  logPath: /var/log/grfplugin_logs

Other parameters have default values. Optionally, you can modify them according to your requirements. To know more about each parameter, see Plugin configuration parameters.

Update the following parameters in the plugin configuration file:
- lbHost: IP address of the cluster node where the GRF plugin will be installed for one node cluster configuration. For three node cluster configuration, (haDeploymentEnabled is true), lbHost is the virtual IP address configured in Nginx and Keepalived.
- lbPort: Default port is 30081 for one node cluster configuration. For three node cluster configuration, (haDeploymentEnabled is true), lbport must be assigned with a free port (preferably above 45000) available on all cluster nodes. This port is used as nginx proxy port for the plugin.
  
  Note: The lbport is used as proxy port for eventlistenernodeport, which is used for subscribing to events.
- grfPluginrootServiceUUID: RootServiceUUID to be used by the GRF plugin service.
Other parameters can have default values. Optionally, you can update them with values based on your requirements. For more information on each parameter, see Plugin configuration parameters.
Generate the Helm package for the GRF plugin on the deployment node.
1. Navigate to odim-controller/helmcharts/grfplugin.
```
cd ~/ODIM/odim-controller/helmcharts/grfplugin
```
2. Run the following command to create grfplugin Helm package at ~/plugins/grfplugin:
```
helm package grfplugin -d ~/plugins/grfplugin
```
  The Helm package for the GRF plugin is created in the tgz format.
Save the GRF plugin Docker image on the deployment node at ~/plugins/grfplugin.
```
docker save grfplugin:1.0 -o ~/plugins/grfplugin/grfplugin.tar
```
Save the proxy configuration file install/templates/grfplugin_proxy_server.conf.j2 to ~/plugins/grfplugin.

Important: Do NOT change the value of any parameter in this file.
Navigate to the /ODIM/odim-controller/scripts directory on the deployment node.
```
 cd ~/ODIM/odim-controller/scripts
```
Open the kube_deploy_nodes.yaml file to edit.
```
vi kube_deploy_nodes.yaml
```

Update the following parameters in the kube_deploy_nodes.yaml file to their corresponding values:

Parameter	Value
connectionMethodConf	The connection method associated with the GRF plugin: ConnectionMethodVariant: `Compute:BasicAuth:GRF_v1.0.0` Check if it is there already before updating. If yes, do not add it again.
odimraKafkaClientCertFQDNSan	The FQDN to be included in the Kafka client certificate of Resource Aggregator for ODIM for deploying the GRF plugin:grfplugin, grfplugin-events Add these values to the existing comma-separated list.
odimraServerCertFQDNSan	The FQDN to be included in the server certificate of Resource Aggregator for ODIM for deploying the GRF plugin: grfplugin, grfplugin-events. Add these values to the existing comma-separated list.
odimPluginPath	The path of the directory where the GRF Helm package, the `grfplugin` image, and the modified `grfplugin-config.yaml` are copied.

Example:

    connectionMethodConf:
      ConnectionMethodType: Redfish
      ConnectionMethodVariant: Compute:BasicAuth:GRF_v1.0.0
    odimraKafkaClientCertFQDNSan: grfplugin,grfplugin-events
    odimraServerCertFQDNSan: grfplugin,grfplugin-events

Move odimra_kafka_client.key, odimra_kafka_client.crt, odimra_server.key and odimra_server.crt stored in odimCertsPath to a different folder.

NOTE: odimCertsPath is the absolute path of the directory where certificates required by the services of Resource Aggregator for ODIM are present. Refer to the "Odim-controller configuration parameters" section in this document for more information on odimCertsPath.

Update odimra-secrets:

python3 odim-controller.py --config /home/${USER}/ODIM/odim-controller/scripts/kube_deploy_nodes.yaml --upgrade odimra-secret

Run the following command:

python3 odim-controller.py --config \ 
/home/${USER}/ODIM/odim-controller/scripts\
/kube_deploy_nodes.yaml --upgrade odimra-config

Run the following command to install the GRF plugin:

python3 odim-controller.py --config \
/home/${USER}/ODIM/odim-controller/scripts\
/kube_deploy_nodes.yaml --add plugin --plugin grfplugin

Run the following command on the cluster nodes to verify the GRF plugin pod is up and running:
```
kubectl get pods -n odim
```
Example output showing the GRF plugin pod details:
```
NAME READY STATUS RESTARTS AGE
grfplugin-5fc4b6788-2xx97 1/1 Running 0 4d22h
```
Navigate to ~/ODIM/odim-controller/scripts.
```
cd ~/ODIM/odim-controller/scripts
```
Add the GRF plugin into the Resource Aggregator for ODIM framework.

Replacing an unreachable controller node with a new one

Set the following environment variables on the deployment node:

export NEW_CONTROLLER_NODE_IP=<IP_address_of_new_controller_node>

export NEW_CONTROLLER_NODE_HOSTNAME=<Hostname_of_new_controller_node>

export DEPLOYMENT_ID=<Deployment_ID_of_the_cluster_being_updated>

export ODIM_CONTROLLER_SRC_PATH=/home/${USER}/ODIM/odim-controller

export ODIM_CONTROLLER_CONFIG_FILE=/home/${USER}/ODIM/\
odim-controller/scripts/kube_deploy_nodes.yaml

export K8S_INVENTORY_FILE=${ODIM_CONTROLLER_SRC_PATH}/kubespray/inventory/k8s-cluster-${DEPLOYMENT_ID}/hosts.yaml

Replace {ODIM_CONTROLLER_SRC_PATH} with:

/home/$\{USER\}/ODIM/odim-controller.

Replace {DEPLOYMENT_ID} with the deployment Id of the cluster being updated.

Perform the following steps on one of the controller nodes:
1. Mark the failed controller node as unschedulable:
```
kubectl cordon <Failed_Controller_HostName>
```
2. Delete the failed controller node from the cluster:
```
kubectl delete node <Failed_Controller_HostName>
```
3. Get the container name of etcd:
```
sudo docker ps | grep etcd | awk '{print $10}'
```
4. List the etcd members to obtain the member Id of the failed controller node.
```
sudo docker exec -it <etcd_container_name> etcdctl member list
```
5. Remove the failed controller node from the etcd cluster:
```
sudo docker exec -it <etcd_container_name> etcdctl member remove <failed_node_etcd_member_id>
```

Perform the following steps on the deployment node:

Enable passwordless login for the new controller node.

/usr/bin/ssh-copy-id -o StrictHostKeyChecking=no -i ~/.ssh/id_rsa.pub ${USER}@${New_Controller_Node_IP}

Verify that time on the new controller node is in sync with the deployment node and other cluster nodes.
Edit ${K8S_INVENTORY_FILE} to:
- Remove the failed controller node details.
- Add the new controller node details under the following sections:
  - etcd
  - kube-master
  - kube-node
  - hosts
```
vi ${K8S_INVENTORY_FILE}
```
Edit \{ODIM\_CONTROLLER\_SRC\_PATH\}/kubespray/inventory/k8s-cluster-$\{DEPLOYMENT\_ID\}/group\_vars/all/all.yml to:
- Update the no_proxy parameter with the new controller node IP.
- Remove the failed controller node IP from the no_proxy list.

Run the following commands:

cp ${ODIM_CONTROLLER_CONFIG_FILE} ${ODIM_CONTROLLER_SRC_PATH}/\
odimra/roles/k8-copy-image/files/helm_config_values.yaml

cd ${ODIM_CONTROLLER_SRC_PATH}/odimra

ansible-playbook -K -i ${K8S_INVENTORY_FILE} --become \
--become-user=root --extra-vars "host=${NEW_CONTROLLER_NODE_HOSTNAME}" \
k8_copy_image.yaml

When prompted for password, enter the sudo password of the node.

cd ${ODIM_CONTROLLER_SRC_PATH}/kubespray

ansible-playbook -K -i ${K8S_INVENTORY_FILE} \
--become --become-user=root --limit=etcd,kube-master \
-e ignore_assert_errors=yes cluster.yml

Run the following command on any of the cluster nodes:

kubectl get nodes -o wide

If any of the nodes are listed as "Ready,Unschedulable", run the following commands on any of the existing controller nodes:

kubectl uncordon <unschedulable_controller_node_name>

cd ${ODIM_CONTROLLER_SRC_PATH}/odimra

cp ${ODIM_CONTROLLER_CONFIG_FILE} ${ODIM_CONTROLLER_SRC_PATH}\
/odimra/roles/pre-install/files/helmcharts/helm_config_values.yaml

cp ${ODIM_CONTROLLER_CONFIG_FILE} ${ODIM_CONTROLLER_SRC_PATH}\
/odimra/roles/odimra-copy-image/files/odimra_config_values.yaml

ansible-playbook -K -i ${K8S_INVENTORY_FILE} \
--become --become-user=root --extra-vars \
"host=${New_Controller_Node_HostName}" pre_install.yaml

rm -f  ${ODIM_CONTROLLER_SRC_PATH}/odimra/roles/\
k8-copy-image/files/helm_config_values.yaml \
${ODIM_CONTROLLER_SRC_PATH}/odimra/roles/pre-install/\
files/helmcharts/helm_config_values.yaml ${ODIM_CONTROLLER_SRC_PATH}/\
odimra/roles/odimra-copy-image/files/odimra_config_values.yaml

[Optional] If the failed node becomes accessible later, run the following commands on the failed node to uninstall Kubernetes:

kubectl drain <node name> --delete-local-data \
--force --ignore-daemonsets

sudo kubeadm reset

sudo apt-get -y autoremove --purge \
kubeadm kubectl kubelet kubernetes-cni

sudo rm -rf /etc/cni /etc/kubernetes /var/lib/dockershim \
/var/lib/etcd /var/lib/kubelet /var/run/kubernetes ~/.kube/*

sudo systemctl restart docker

Replacing an unreachable controller node with an existing worker node

Set the following environment variables on the deployment node:

export EXISTING_WORKER_NODE_IP=<IP_address_of_existing_worker_node>

export EXISTING_WORKER_NODE_HOSTNAME=<Hostname_of_existing_worker_node>

export DEPLOYMENT_ID=<Deployment_ID_of_the_cluster_being_updated>

export ODIM_CONTROLLER_SRC_PATH=/home/${USER}/ODIM/odim-controller

export ODIM_CONTROLLER_CONFIG_FILE=/home/${USER}/ODIM/odim-controller/\
scripts/kube_deploy_nodes.yaml

export K8S_INVENTORY_FILE=${ODIM_CONTROLLER_SRC_PATH}/kubespray/inventory/k8s-cluster-${DEPLOYMENT_ID}/hosts.yaml

Replace \{ODIM\_CONTROLLER\_SRC\_PATH\} with:

/home/$\{USER\}/ODIM/odim-controller.

Replace {DEPLOYMENT_ID} with the deployment Id of the cluster being updated.

Perform the following steps on one of the controller nodes:
1. Mark the failed controller node as unschedulable:
```
kubectl cordon <Failed_Controller_HostName>
```
2. Delete the failed controller node from the cluster:
```
kubectl delete node <Failed_Controller_HostName>
```
3. Get the container name of etcd:
```
sudo docker ps | grep etcd | awk '{print $10}'
```
4. List the etcd members to obtain the member Id of the failed controller node.
```
sudo docker exec -it <etcd_container_name> etcdctl member list
```
5. Remove the failed controller node from the etcd cluster:
```
sudo docker exec -it <etcd_container_name> etcdctl member remove <failed_node_etcd_member_id>
```

Perform the following steps on the deployment node:

Remove the existing worker node. To know how to remove a node, see step 1 in Scaling down the resources and services of HPE Resource Aggregator for ODIM.
Edit $\{K8S\_INVENTORY\_FILE\} to add the removed worker node as a new controller node with required details under the following sections.
- etcd
- kube-master
- kube-node
- hosts
```
vi ${K8S_INVENTORY_FILE}
```
Edit \{ODIM\_CONTROLLER\_SRC\_PATH\}/kubespray/inventory/k8s-cluster-$\{DEPLOYMENT\_ID\}/group\_vars/all/all.yml to:
- Update the no_proxy parameter with the removed worker node IP.
- Remove the failed controller node IP from the no_proxy list.

Run the following commands:

cp ${ODIM_CONTROLLER_CONFIG_FILE} ${ODIM_CONTROLLER_SRC_PATH}/\
odimra/roles/k8-copy-image/files/helm_config_values.yaml

cd ${ODIM_CONTROLLER_SRC_PATH}/odimra

ansible-playbook -K -i ${K8S_INVENTORY_FILE} --become \
--become-user=root --extra-vars "host=${EXISTING_WORKER_NODE_HOSTNAME}" \
k8_copy_image.yaml

When prompted for password, enter the sudo password of the node.

cd ${ODIM_CONTROLLER_SRC_PATH}/kubespray

ansible-playbook -K -i ${K8S_INVENTORY_FILE} \
--become --become-user=root --limit=etcd,kube-master \
-e ignore_assert_errors=yes cluster.yml

Run the following command on any of the cluster nodes:

kubectl get nodes -o wide

If any of the nodes are listed as "Ready,Unschedulable", run the following commands on any of the existing controller nodes:

kubectl uncordon <unschedulable_controller_node_name>

cd ${ODIM_CONTROLLER_SRC_PATH}/odimra

cp ${ODIM_CONTROLLER_CONFIG_FILE} ${ODIM_CONTROLLER_SRC_PATH}\
/odimra/roles/pre-install/files/helmcharts/helm_config_values.yaml

cp ${ODIM_CONTROLLER_CONFIG_FILE} ${ODIM_CONTROLLER_SRC_PATH}\
/odimra/roles/odimra-copy-image/files/odimra_config_values.yaml

ansible-playbook -K -i ${K8S_INVENTORY_FILE} \
--become --become-user=root --extra-vars \
"host=${EXISTING_WORKER_NODE_HOSTNAME}" pre_install.yaml

rm -f  ${ODIM_CONTROLLER_SRC_PATH}/odimra/roles/\
k8-copy-image/files/helm_config_values.yaml \
${ODIM_CONTROLLER_SRC_PATH}/odimra/roles/pre-install/\
files/helmcharts/helm_config_values.yaml ${ODIM_CONTROLLER_SRC_PATH}/\
odimra/roles/odimra-copy-image/files/odimra_config_values.yaml

[Optional] If the failed node becomes accessible later, uninstall Kubernetes from it:

kubectl drain <node name> --delete-local-data \
--force --ignore-daemonsets

sudo kubeadm reset

sudo apt-get -y autoremove --purge \
kubeadm kubectl kubelet kubernetes-cni

sudo rm -rf /etc/cni /etc/kubernetes /var/lib/dockershim \
/var/lib/etcd /var/lib/kubelet /var/run/kubernetes ~/.kube/*

sudo systemctl restart docker

Removing an existing plugin

To remove an existing plugin from the Resource Aggregator for ODIM framework, run the following command:

python3 odim-controller.py --config \
/home/${USER}/ODIM/odim-controller/scripts/kube_deploy_nodes.yaml \
--remove plugin --plugin <plugin_name>

Uninstalling the resource aggregator services

To remove all the resource aggregator services, run the following command:

python3 odim-controller.py --reset odimra --config \
/home/${USER}/ODIM/odim-controller/scripts/kube_deploy_nodes.yaml

To uninstall all the resource aggregator services by ignoring any errors you may encounter, use the following command:

python3 odim-controller.py --reset odimra --config \
/home/${USER}/ODIM/odim-controller/scripts/kube_deploy_nodes.yaml \
--ignore-errors

CI process

GitHub action workflows, also known as checks, are added to the ODIM repository. They are triggered whenever a Pull Request (PR) is raised against the master (development) branch. The result from the workflow execution is then updated to the PR.

NOTE: You can review and merge PRs only if the checks are passed.

Following checks are added as part of the CI process:

Sl No.	Workflow Name	Description
1	`build_unittest.yml`	Builds and runs Unit Tests with code coverage enabled.
2	`build_deploy_test.yml`	Builds, deploys, runs sanity tests, and uploads build artifacts (like odimra logs).

These checks run in parallel and take approximately 9 minutes to complete.

GitHub action workflow details

build_unittest.yml
- Brings up a Ubuntu 18.04 VM hosted on GitHub infrastructure with preinstalled packages mentioned in the link: https://github.com/actions/virtual-environments/blob/master/images/linux/Ubuntu1804-README.md
- Installs Go 1.13.8 package
- Installs and configures Redis 5.0.8 with two instances running on ports 6379 and 6380
- Checks out the PR code into the Go module directory
- Builds/compiles the code
- Runs the unit tests
build_deploy_test.yml
- Brings up a Ubuntu 18.04 VM hosted on GitHub infrastructure with preinstalled packages mentioned in the link: https://github.com/actions/virtual-environments/blob/master/images/linux/Ubuntu1804-README.md
- Checks out the PR code
- Builds and deploys the following docker containers:
  - ODIMRA
  - Generic Redfish plugin
  - Unmanaged Rack Plugin
  - Kakfa
  - Zookeeper
  - etcd
  - Redis
- Runs the sanity tests
- Prepares build artifacts
- Uploads the build artifacts

NOTE: Build status notifications having a link to the GitHub Actions build job page will be sent to the developer’s email address.

Screenshots of the checks after execution

screenshot

Directories ¶

Path	Synopsis
odim-controller
scripts

?	: This menu
/	: Search site
f or F	: Jump to
y or Y	: Canonical URL

README ¶

Table of contents

Introduction

Filing Resource Aggregator for ODIM defects

Resource Aggregator for ODIM deployment overview

Deployment considerations

Resource Aggregator for ODIM compatibility matrix

Predeployment procedures

Setting up the environment

Pulling the Docker images of all the Kubernetes microservices

Building Docker images of all the services

Generating an encrypted node password

Log path for odim-controller

Deploying Resource Aggregator for ODIM and the Plugins

Deploying the resource aggregator services

Deploying the Unmanaged Rack Plugin

Deploying the Dell plugin

Deploying the Cisco ACI plugin

Adding a plugin into the Resource Aggregator for ODIM framework

Use cases for Resource Aggregator for ODIM

Adding a server into the resource inventory

Viewing the resource inventory

Configuring BIOS settings for a server

Resetting a server

Setting one time boot path for a server

Searching the inventory for specific servers

Updating software and firmware

Subscribing to southbound events

Viewing network fabrics

Creating and deleting volumes

Removing a server from the resource inventory

Using odim-controller command-line interface

Postdeployment operations

Scaling up the resources and services of Resource Aggregator for ODIM

Scaling down the resources and services of Resource Aggregator for ODIM

Rolling back to an earlier deployment revision

Upgrading the Resource Aggregator for ODIM deployment

Appendix

Setting proxy configuration

Setting up time sync across nodes

Downloading and installing Go

Configuring Docker proxy

Installing Docker

Odim-controller configuration parameters

Running curl commands on a different server

Resource Aggregator for ODIM deployment names

Using your own CA certificates and keys

Regenerating certificates

Updating Kafka password and certificate

Updating Zookeeper password and certificate

Updating certificates with SAN entries

Updating /etc/hosts in the containers

Appending CA certificates to the existing Root CA certificate

Resource Aggregator for ODIM default ports

Deploying the GRF plugin

Replacing an unreachable controller node with a new one

Replacing an unreachable controller node with an existing worker node

Removing an existing plugin

Uninstalling the resource aggregator services

CI process

GitHub action workflow details

Screenshots of the checks after execution

Directories ¶

Updating `/etc/hosts` in the containers