QuickStart Guide

Install / Upgrade Manufacturing Connect

34min

For upgrading Manufacturing Connect (MC), refer to the "Upgrading Manufacturing Connect" section below (at the end of this article) for details. 
You can install the Manufacturing Connect in the Google Cloud Platform (GCP).
This process includes the following procedures:
Pre-installation Tasks
Prepare GCP Project
Deploy The Manufacturing Connect through the Google Cloud Marketplace
Post-installation Tasks
Set Up Google Authentication (optional)
Before You Begin
For Google Cloud Platform (GCP): Ensure you have access to a project (with respective project ID) for the Manufacturing Connect instance.
For Google Cloud Platform (GCP): Ensure you have a zone and a respective region for the Manufacturing Connect instance to use.
(Optional) Install Google Manufacturing Data Engine (MDE) to simplify some of the following incoming installation steps.
Step 1: Pre-deployment Tasks
You can create the necessary GCP infrastructure and deploy the Manufacturing Connect dependencies step-by-step.
Step 1a: Access the Google Cloud Shell
You will first need to access the Google Cloud Shell to configure the environment. 
To access the Google Cloud Shell:
﻿Access Google Cloud Shell or install gcloud-cli using the instructions within Install the gcloud CLI﻿
Launch Google Cloud Shell or a local command-line terminal gcloud-cli to execute the following incoming commands. Note: When asked for any type of permissions, always click Enable.
Refer to the following commands to set up the environment.
Step 1b: Define Initial Environment Variables 
Use the following commands to define the initial environment variables. 
Linux
export GCP_PROJECT_ID='your-gcp-project-id'
export GCP_PROJECT_ID='your-gcp-project-id'
﻿
Linux
export REGION='us-central1'
export REGION='us-central1'
﻿
Linux
export ZONE='us-central1-c'
export ZONE='us-central1-c'
﻿
Linux
export NETWORK_NAME="sfp-private-network"
export NETWORK_NAME="sfp-private-network"
﻿
Linux
export SUBNET_NAME="sfp-subnet"
export SUBNET_NAME="sfp-subnet"
﻿

Linux
export GKECLUSTER="mc-cluster"
export GKECLUSTER="mc-cluster"
﻿
Step 1c: Enable GCP Services
Use the following command to enable GCP services. 
Linux
gcloud services enable --project=${GCP_PROJECT_ID} container.googleapis.com
gcloud services enable --project=${GCP_PROJECT_ID} container.googleapis.com
﻿
Step 1d: (Optional) Install Network
When Manufacturing Data Engine (MDE) is not installed, a network can be instead installed by entering the following command. 
Note: Installing Manufacturing Data Engine (MDE) automatically sets up a network.
Linux
gcloud compute networks create "$NETWORK_NAME" --project="$GCP_PROJECT_ID" \
							--description="MDE private network" \
							--subnet-mode=custom \
							--mtu=1460 \
						--bgp-routing-mode=regional
gcloud compute networks create "$NETWORK_NAME" --project="$GCP_PROJECT_ID" \
							--description="MDE private network" \
							--subnet-mode=custom \
							--mtu=1460 \
						--bgp-routing-mode=regional
﻿
Linux
gcloud compute networks subnets create "$SUBNET_NAME" --project="$GCP_PROJECT_ID" \
							--range=10.154.0.0/20 \
							--network="$NETWORK_NAME" \
							--region="$REGION" \
						--enable-private-ip-google-access
gcloud compute networks subnets create "$SUBNET_NAME" --project="$GCP_PROJECT_ID" \
							--range=10.154.0.0/20 \
							--network="$NETWORK_NAME" \
							--region="$REGION" \
						--enable-private-ip-google-access
﻿
Step 1e: Create GKE Cluster
To create a GKE cluster, enter one of the two GKE Cluster Deployment methods:
Important: 
Manufacturing Connect's services are deployed into a Kubernetes cluster. The cluster must be created before deploying the Manufacturing Connect from the Google Cloud Marketplace. UBUNTU_CONTAINERD is the only supported image type for GKE nodes. The Manufacturing Connect can be either deployed into a public or private GKE cluster.
The cluster network where Manufacturing Connect(MC) and Google Manufacturing Data Engine (MDE) are located should be the same to ensure communication between the two. 
Public GKE Cluster Deployment  
To deploy Manufacturing Connect to a public GKE Cluster, enter the following command. 
Linux
gcloud beta container --project "${GCP_PROJECT_ID}" clusters create $GKECLUSTER --zone "${ZONE}" \
							--no-enable-basic-auth --release-channel "regular" --machine-type "e2-standard-2" --image-type "UBUNTU_CONTAINERD" \
							--disk-type "pd-standard" --disk-size "100" --metadata disable-legacy-endpoints=true \
							--scopes "https://www.googleapis.com/auth/devstorage.read_only","https://www.googleapis.com/auth/logging.write","https://www.googleapis.com/auth/monitoring","https://www.googleapis.com/auth/servicecontrol","https://www.googleapis.com/auth/service.management.readonly","https://www.googleapis.com/auth/trace.append" \
							--max-pods-per-node "110" --num-nodes "3" --logging=SYSTEM,WORKLOAD --monitoring=SYSTEM \
							--enable-ip-alias  --no-enable-intra-node-visibility --default-max-pods-per-node "110" \
							--no-enable-master-authorized-networks --addons HorizontalPodAutoscaling,HttpLoadBalancing,GcePersistentDiskCsiDriver \
							--enable-autoupgrade --enable-autorepair --max-surge-upgrade 1 --max-unavailable-upgrade 0 \
							--maintenance-window-start "2022-05-21T02:00:00Z" --maintenance-window-end "2022-05-22T02:00:00Z" \
							--maintenance-window-recurrence "FREQ=WEEKLY;BYDAY=MO,TU,WE,TH,FR,SA,SU" \
							--workload-pool "${GCP_PROJECT_ID}.svc.id.goog" --enable-shielded-nodes --node-locations "${ZONE}" \
						--network "${NETWORK_NAME}" --subnetwork "${SUBNET_NAME}" --labels "goog-packaged-solution=mfg-mde"
gcloud beta container --project "${GCP_PROJECT_ID}" clusters create $GKECLUSTER --zone "${ZONE}" \
							--no-enable-basic-auth --release-channel "regular" --machine-type "e2-standard-2" --image-type "UBUNTU_CONTAINERD" \
							--disk-type "pd-standard" --disk-size "100" --metadata disable-legacy-endpoints=true \
							--scopes "https://www.googleapis.com/auth/devstorage.read_only","https://www.googleapis.com/auth/logging.write","https://www.googleapis.com/auth/monitoring","https://www.googleapis.com/auth/servicecontrol","https://www.googleapis.com/auth/service.management.readonly","https://www.googleapis.com/auth/trace.append" \
							--max-pods-per-node "110" --num-nodes "3" --logging=SYSTEM,WORKLOAD --monitoring=SYSTEM \
							--enable-ip-alias  --no-enable-intra-node-visibility --default-max-pods-per-node "110" \
							--no-enable-master-authorized-networks --addons HorizontalPodAutoscaling,HttpLoadBalancing,GcePersistentDiskCsiDriver \
							--enable-autoupgrade --enable-autorepair --max-surge-upgrade 1 --max-unavailable-upgrade 0 \
							--maintenance-window-start "2022-05-21T02:00:00Z" --maintenance-window-end "2022-05-22T02:00:00Z" \
							--maintenance-window-recurrence "FREQ=WEEKLY;BYDAY=MO,TU,WE,TH,FR,SA,SU" \
							--workload-pool "${GCP_PROJECT_ID}.svc.id.goog" --enable-shielded-nodes --node-locations "${ZONE}" \
						--network "${NETWORK_NAME}" --subnetwork "${SUBNET_NAME}" --labels "goog-packaged-solution=mfg-mde"
﻿
Private GKE Cluster Deployment
To deploy Manufacturing Connect to a private GKE Cluster, enter the following command. Modify Kubernetes control nodes CIDR --master-ipv4-cidr 10.155.1.0/28 for your network setup.
Linux
gcloud beta container clusters create $GKECLUSTER --project "$GCP_PROJECT_ID" --zone "$ZONE" \
            --no-enable-basic-auth --release-channel "regular" --machine-type "e2-standard-2" --image-type "UBUNTU_CONTAINERD" \
            --disk-type "pd-standard" --disk-size "100" --metadata disable-legacy-endpoints=true \
            --scopes "https://www.googleapis.com/auth/devstorage.read_only","https://www.googleapis.com/auth/logging.write","https://www.googleapis.com/auth/monitoring","https://www.googleapis.com/auth/servicecontrol","https://www.googleapis.com/auth/service.management.readonly","https://www.googleapis.com/auth/trace.append" \
            --max-pods-per-node "110" --num-nodes "3" --logging=SYSTEM,WORKLOAD --monitoring=SYSTEM \
            --enable-ip-alias  --no-enable-intra-node-visibility --default-max-pods-per-node "110" \
            --enable-master-authorized-networks --addons HorizontalPodAutoscaling,HttpLoadBalancing,GcePersistentDiskCsiDriver \
            --enable-autoupgrade --enable-autorepair --max-surge-upgrade 1 --max-unavailable-upgrade 0 \
            --maintenance-window-start "2022-11-21T02:00:00Z" --maintenance-window-end "2022-11-22T02:00:00Z" \
            --maintenance-window-recurrence "FREQ=WEEKLY;BYDAY=MO,TU,WE,TH,FR,SA,SU" \
            --workload-pool "$GCP_PROJECT_ID.svc.id.goog" --enable-shielded-nodes --node-locations "$ZONE" \
            --network "$NETWORK_NAME" --subnetwork "$SUBNET_NAME" --labels "goog-packaged-solution=mfg-mde" \
            --enable-private-nodes --enable-private-endpoint --master-ipv4-cidr 10.155.1.0/28
gcloud beta container clusters create $GKECLUSTER --project "$GCP_PROJECT_ID" --zone "$ZONE" \
            --no-enable-basic-auth --release-channel "regular" --machine-type "e2-standard-2" --image-type "UBUNTU_CONTAINERD" \
            --disk-type "pd-standard" --disk-size "100" --metadata disable-legacy-endpoints=true \
            --scopes "https://www.googleapis.com/auth/devstorage.read_only","https://www.googleapis.com/auth/logging.write","https://www.googleapis.com/auth/monitoring","https://www.googleapis.com/auth/servicecontrol","https://www.googleapis.com/auth/service.management.readonly","https://www.googleapis.com/auth/trace.append" \
            --max-pods-per-node "110" --num-nodes "3" --logging=SYSTEM,WORKLOAD --monitoring=SYSTEM \
            --enable-ip-alias  --no-enable-intra-node-visibility --default-max-pods-per-node "110" \
            --enable-master-authorized-networks --addons HorizontalPodAutoscaling,HttpLoadBalancing,GcePersistentDiskCsiDriver \
            --enable-autoupgrade --enable-autorepair --max-surge-upgrade 1 --max-unavailable-upgrade 0 \
            --maintenance-window-start "2022-11-21T02:00:00Z" --maintenance-window-end "2022-11-22T02:00:00Z" \
            --maintenance-window-recurrence "FREQ=WEEKLY;BYDAY=MO,TU,WE,TH,FR,SA,SU" \
            --workload-pool "$GCP_PROJECT_ID.svc.id.goog" --enable-shielded-nodes --node-locations "$ZONE" \
            --network "$NETWORK_NAME" --subnetwork "$SUBNET_NAME" --labels "goog-packaged-solution=mfg-mde" \
            --enable-private-nodes --enable-private-endpoint --master-ipv4-cidr 10.155.1.0/28
﻿
Step 2: Prepare GCP Project
Follow this guide: Managing billing for Cloud Marketplace products﻿
Step 3: Deploy Manufacturing Connect through the Google Cloud Marketplace
To purchase the Manufacturing Connect for a billing account:
Using a Google Cloud Provider Billing Admin role, open Manufacturing Connect from the Google Marketplace page.
Click PURCHASE.
Click the MANAGE ACCOUNTS button.
Link the service account that was created in the previous step: Prepare GCP project.
It is advisable to create the service account in the same project where Manufacturing Connect will be deployed.
Deleting the service account will disrupt GCP Marketplace Billing and may remove the product license.
To deploy the Manufacturing Connect to a particular GCP project:
After you've created a Kubernetes cluster, you can deploy the Manufacturing Connect from the Google Cloud Marketplace.﻿
Log into the Google Cloud Marketplace.
Click the Configure button.
In dropdown list Reporting service account, select the service account name that was created in the previous step: Prepare GCP Project.
If the Manufacturing Connect is deployed to a private GKE cluster, then select the Internal Load Balancer option.
Refer to the following table to learn more about GCP parameters. 
Parameter
Description
Existing Kubernetes Cluster
Select the name of the GKE Cluster that was deployed in Step 1: Pre-installation Tasks. You can also select Or Create New Cluster to specify a different GKE cluster, if needed.
Namespace
This is the namespace where the Manufacturing Connect will be deployed to. You cannot have two identical namespaces in the same Google Kubernetes Engine (GKE) cluster.
Create a new namespace. See Kubernetes Namespaces for more details. 
App instance name
This is the application name of the Manufacturing Connect instance. In most situations, it can be left as the default value intelligent-manufacturing-con-1. However, if you have multiple Manufacturing Connect instances or multiple applications in your GKE cluster, changing this value will help you uniquely identify each instance.
While you could have identical names, this is not recommended to prevent confusion.
Use Internal TCP/UDP Load Balancer
If the Manufacturing Connect will be deployed to a private GKE cluster, Select Yes. Select No otherwise.
When Yes is selected, then Frontend Load Balancer static IP address and Remote Access Load Balancer static IP Address will only be accessible within the Virtual Private Cloud (VPC) network where the private GKE cluster is located. These two addresses will not be connected to the internet.
Frontend Load Balancer static IP address
This is the load balancer for TCP-based protocols such as HTTPS and MQTT. When specified, it will provide a consistent endpoint for Manufacturing Connect Edge to connect to. This eliminates the need for Manufacturing Connect Edge reactivation in the case of Manufacturing Connect redeployment. Both Frontend Load Balancer static IP address and Remote Access Load Balancer static IP address must be specified to eliminate the need for Manufacturing Connect Edge reactivation.
This value cannot be identical to Remote Access Load Balancer static IP Address.
When an IP address is not specified, Manufacturing Connect will automatically specify a dynamic IP address for this parameter instead. If Manufacturing Connect has to be redeployed, you can obtain the dynamic IP address and specify it as the new Frontend Load Balancer static IP address. Otherwise, any Manufacturing Connect Edge connected to the redeployed Manufacturing Connect will have to be reactivated﻿. 
Remote Access Load Balancer static IP Address
This is the load balancer for UDP-based protocols. When specified, it will provide a consistent endpoint for Manufacturing Connect Edge to connect to. This eliminates the need for Manufacturing Connect Edge reactivation in the case of Manufacturing Connect redeployment. Both Frontend Load Balancer static IP address and  Remote Access Load Balancer static IP address must be specified to eliminate the need for Manufacturing Connect Edge reactivation.
This value cannot be identical to Frontend Load Balancer static IP address.
When an IP address is not specified, Manufacturing Connect will automatically specify a dynamic IP address for this parameter instead. If Manufacturing Connect has to be redeployed, you can obtain the dynamic IP address and specify it as the new Remote Access Load Balancer static IP Address. Otherwise, any Manufacturing Connect Edge connected to the redeployed Manufacturing Connect will have to be reactivated﻿.  
RemoteAccess network for Edge devices
This is the network range used for communication between Manufacturing Connect and Manufacturing Connect Edge. By default, this range is 192.168.127.0/24.
If your internal network uses this same range, then there is a potential for conflict. You may not be able to access a Manufacturing Connect Edge remotely from Manufacturing Connect.  Likewise, if you have third-party software (for example, an MQTT server) located in this IP range, your Manufacturing Connect Edge will not be able to access the software (in this case, fail to connect to the MQTT server).
Specify a different range to avoid IP address conflict. 
MDE integration topic
By default, this is input-messages. When you deploy MDE (and the Pub/Sub topic was changed), change this value to the equivalent MDE deployment Pub/Sub topic.   
Create new service account
Select the service account name that was created in Step 2: Prepare GCP Project.
After the project is successfully deployed, you will receive an IP address on where the Manufacturing Connect is located.
Note: After Manufacturing Connect is deployed, it will automatically attempt to Activate the License Server﻿ and receive a site license from the license server. This site license will be later used to activate Manufacturing Connect edge instances that use the site-license. If the Manufacturing Connect is located on a private network, then you must Activate the License Server﻿ manually.
Step 4: Post-deployment Tasks
Refer to the following post-deployment tasks. 
Verify Site License Installation
Log in to Manufacturing Connect Admin Console. See the Access the Admin Console section of Access to Manufacturing Connect﻿ for details.
From the Navigation panel, select License Server.
The License Server Management pane appears. 
From Core license list, verify there is a license. If no license exists, you must Activate the License Server﻿. 
MC Admin Console : License Server:Core license installed
﻿
Change Initial Credentials (Mandatory)
Open GKE applications list in your GCP project.
Click the Kubernetes application which was just installed.
Click the Show Info Panel.
From the Show Info Panel, follow instructions to get the application URLs and initial credentials.
Change the MC Admin password.
Change the MC Keycloak Admin password.
Upload GCP Credentials (Optional)
Open the MC Admin Console.
Click Settings/Cloud Settings.
Follow instructions for Generate Key in the Cloud Credentials section.
Set GCS Bucket (Optional)
Open MC Admin Console.
Click Settings/Cloud Settings.
Select Google Cloud Storage in Storage settings.
To create a bucket, follow instructions for Create bucket.
Set bucket name.
Click Save.
Associate a Domain Name with MC Instance
This is mandatory only if Google Authorization is needed for the MC instance.
Associate MC external IP address with a domain name.
Open the MC Admin Console.
Open Settings/Entrypoints.
Set the new domain name.
Click Save.
Step 5: Setup Google Authentication (optional)
If Google Authentication is desired for the Manufacturing Connect instance, a domain name must be associated with the Manufacturing Connect instance.
OAuth Consent Screen
Open the APIs & Services OAuth consent screen.
Select External type.
Note: In testing mode, External allows add up to 100 Google accounts from any organization. Internal type allows only users within the current organization.
     3. On the next screen set mandatory attributes.
App Name
User support email
Add authorized domain
Example: If for Manufacturing Connect we use domain name test.mc.domain.com, then use domain.com as an authorized domain.
Developer contact information
Create OAuth 2.0 Credentials
Open the APIs & Services OAuth credentials screen.
Create OAuth client ID with the following field-value pairs.
Application type: Web application
Authorized redirect url: https://<your-domain>/auth/realms/standalone/broker/google/endpoint
Set up Manufacturing Connect Keycloak
Open the Keycloak Admin console.
Click Identity Providers.
Select Google from the list
Set Client ID and Secret ID values from the previous step: Create OAuth 2.0 credentials.
Set First Login Flow to google-login.
Click Save.
Grant Permissions to Google Accounts
Open the MC Admin Console.
Click Users.
Add a new user (set Google email).
Enable the user.
Grant Admin role if required.
﻿
Upgrade Manufacturing Connect
Refer to the following steps to upgrade Manufacturing Connect (MC). 
See the Manufacturing Connect installation guide from GitHub to learn more. 
Step 1: Retrieve the git repository
Linux
git clone https://github.com/litmusautomation/mc-gcp-marketplace.git
git clone https://github.com/litmusautomation/mc-gcp-marketplace.git
﻿
Step 2: Enter the retrieved repository
Linux
cd mc-gcp-marketplace
cd mc-gcp-marketplace
﻿
Step 3: Run upgrade script
Linux
./upgrade-mc.sh $GCP_PROJECT_ID $ZONE $GKECLUSTER $NAMESPACE $TARGET_VERSION
./upgrade-mc.sh $GCP_PROJECT_ID $ZONE $GKECLUSTER $NAMESPACE $TARGET_VERSION
﻿
Specify the following parameters:
$GCP_PROJECT_ID is the Google Cloud Platform (GCP) project ID.
$ZONE is the Zone where GKE cluster is deployed.
$GKECLUSTER is the GKE cluster name.
$NAMESPACE is the Kubernetes namespace where Manufacturing Connect is deployed.
$TARGET_VERSION is the required Manufacturing Connect target version, for example 2.8.0-120.
﻿

Parameter	Description
Existing Kubernetes Cluster	Select the name of the GKE Cluster that was deployed in Step 1: Pre-installation Tasks. You can also select Or Create New Cluster to specify a different GKE cluster, if needed.
Namespace	This is the namespace where the Manufacturing Connect will be deployed to. You cannot have two identical namespaces in the same Google Kubernetes Engine (GKE) cluster. Create a new namespace. See Kubernetes Namespaces for more details.
App instance name	This is the application name of the Manufacturing Connect instance. In most situations, it can be left as the default value intelligent-manufacturing-con-1. However, if you have multiple Manufacturing Connect instances or multiple applications in your GKE cluster, changing this value will help you uniquely identify each instance. While you could have identical names, this is not recommended to prevent confusion.
Use Internal TCP/UDP Load Balancer	If the Manufacturing Connect will be deployed to a private GKE cluster, Select Yes. Select No otherwise. When Yes is selected, then Frontend Load Balancer static IP address and Remote Access Load Balancer static IP Address will only be accessible within the Virtual Private Cloud (VPC) network where the private GKE cluster is located. These two addresses will not be connected to the internet.
Frontend Load Balancer static IP address	This is the load balancer for TCP-based protocols such as HTTPS and MQTT. When specified, it will provide a consistent endpoint for Manufacturing Connect Edge to connect to. This eliminates the need for Manufacturing Connect Edge reactivation in the case of Manufacturing Connect redeployment. Both Frontend Load Balancer static IP address and Remote Access Load Balancer static IP address must be specified to eliminate the need for Manufacturing Connect Edge reactivation. This value cannot be identical to Remote Access Load Balancer static IP Address. When an IP address is not specified, Manufacturing Connect will automatically specify a dynamic IP address for this parameter instead. If Manufacturing Connect has to be redeployed, you can obtain the dynamic IP address and specify it as the new Frontend Load Balancer static IP address. Otherwise, any Manufacturing Connect Edge connected to the redeployed Manufacturing Connect will have to be reactivated.
Remote Access Load Balancer static IP Address	This is the load balancer for UDP-based protocols. When specified, it will provide a consistent endpoint for Manufacturing Connect Edge to connect to. This eliminates the need for Manufacturing Connect Edge reactivation in the case of Manufacturing Connect redeployment. Both Frontend Load Balancer static IP address and Remote Access Load Balancer static IP address must be specified to eliminate the need for Manufacturing Connect Edge reactivation. This value cannot be identical to Frontend Load Balancer static IP address. When an IP address is not specified, Manufacturing Connect will automatically specify a dynamic IP address for this parameter instead. If Manufacturing Connect has to be redeployed, you can obtain the dynamic IP address and specify it as the new Remote Access Load Balancer static IP Address. Otherwise, any Manufacturing Connect Edge connected to the redeployed Manufacturing Connect will have to be reactivated.
RemoteAccess network for Edge devices	This is the network range used for communication between Manufacturing Connect and Manufacturing Connect Edge. By default, this range is 192.168.127.0/24. If your internal network uses this same range, then there is a potential for conflict. You may not be able to access a Manufacturing Connect Edge remotely from Manufacturing Connect. Likewise, if you have third-party software (for example, an MQTT server) located in this IP range, your Manufacturing Connect Edge will not be able to access the software (in this case, fail to connect to the MQTT server). Specify a different range to avoid IP address conflict.
MDE integration topic	By default, this is input-messages. When you deploy MDE (and the Pub/Sub topic was changed), change this value to the equivalent MDE deployment Pub/Sub topic.
Create new service account	Select the service account name that was created in Step 2: Prepare GCP Project.

Updated 10 Dec 2024

System Requirements

Access to Manufacturing Connect