Help us improve your experience.

Let us know what you think.

Do you have time for a two-minute survey?

 
ON THIS PAGE
 

Provision AWS Compute and Networking Components

Set up EC2 instances, storage, networking, and load balancing for the cluster.

This topic explains how to create and configure the core AWS infrastructure for the deployment. It includes launching EC2 instances, configuring storage and networking, and setting up target groups and load balancers. These components provide the compute capacity and connectivity required for the Routing Director cluster.

Launch the EC2 VMs

Launch the EC2 virtual machines (VMs). Create the VMs with the following bare minimum requirements:

  • Select the VM flavor—The VM flavor must be at least c5.4xlarge and meet the specified minimum requirements for vCPU and memory described in Hardware Requirements.

  • Configure the network settings:

    • If the VMs do not have public IP addresses assigned, the VMs must run on a subnet with NAT gateway enabled.

    • If the VMs have elastic public IP addresses assigned, the VMs should run on a subnet with internet gateway (IGW) enabled.

  • Configure the following storage specifications:

    • The bare minimum requirement is 400-GB for the primary disk and 100-GB for the Ceph disk.

    • Volume type must be General purpose SSD (gp3).

    • IOPS must be minimum 5000.

    • Throughput must be minimum 1000 MiB/s.

To launch the EC2 VMs.

  1. Create the VMs. You can create each VM one at a time or all four at a time.

    Click EC2 > Instances > Launch instance. Enter information by referring to the following table: For fields not defined in this table, retain the default values.

    Table 1:
    Field Description
    Name and tags

    Name

    Enter a name for the VM. If you are creating all the VMs at once, the same name is used for all VMs. For example, rd-nodes or primary1.

    You can create all VMs at once if all the VMs are in the same subnet. For high availability of nodes, if all the nodes are in different subnets, create each node one at a time.
    Summary

    Number of instances

    Enter 1 if you are creating each VM one at a time, or enter 4 if you are creating all VMs at once.
    Application and OS Images (Amazon Machine Image)

    Click My AMIs.

    Owned by me

    Select the checkbox to display only images imported by you.

    Amazon Machine Image (AMI)

    Select the image with the ImportTaskId and ImageId that you noted down earlier after the converting the image to AMI.

    The OS of the image and the architecture are automatically detected and displayed.
    Instance type

    Instance type

    Select the c5.4xlarge instance type.
    Key pair (login)

    Key pair name - required

    Select Proceed without a key pair to proceed since you can use the VM console to configure the VM instance and do not need to directly SSH to the VM.

    Alternatively, enter a key pair name if you have an existing one and want to directly SSH to the VM instance. Or, you can click Create new key pair to create a new key pair and enter a name for the key pair. If you choose to create a new key pair, ensure that you save the file with the private key information that is automatically downloaded.
    Network settings

    VPC -required

    Select the VPC for the cluster. The VPC must be the same as the one selected when creating security groups.

    Subnet

    Select the subnet that you want to use for the cluster. This subnet is used for all the VM instances you have chosen to create.

    To ensure high availability, if your VMs are in different subnets, you need to configure multiple subnets. To configure multiple subnets, you must create each VM instance one at a time and cannot create all instances at once.

    Auto-assign public IP

    Select Disable if the subnet has NAT gateway enabled. Select Enable if the subnet has IGW enabled.

    Firewall (security groups)

    Click Select existing security group to select the previously created security group.

    Common security groups

    Select the previously created security group. In this example, select EC2-sg.

    Configure storage

    Click Advanced.
    Storage (volumes) > EBS Volumes
    Volume 1 (AMI Root) (Custom)

    Size (GiB)

    Enter the main disk size. The minimum required size is 400-GiB.

    Volume type

    Select General purpose SSD (gp3).

    IOPS

    Enter 5000.

    Delete on termination

    We recommend that you select Yes. When selected, the disk is deleted whenever the VM is deleted.

    Throughput

    Enter 1000.
    Volume 2 (AMI) (Custom)

    Size (GiB)

    Enter the size required for the Ceph disk. The minimum required size is 100-GiB.

    Volume type

    Select General purpose SSD (gp3).

    IOPS

    Enter 5000.

    Delete on termination

    We recommend that you select Yes. When selected, the disk is deleted whenever the VM is deleted.

    Throughput

    Enter 1000.
    Advanced details

    Hostname type

    Select Resource name.
  2. Click Launch instance. The instance is created and listed under EC2 > Instances along with the instance ID. Click ⭮ to refresh the page if the instance is not displayed immediately. If you created multiple instances at once, all instances are created with the same name. The VMs are ready to be provisioned when the Instance state is Running and the Status check has transitioned from Initializing to Completed.
  3. Rename the VMs.

    If you created all the instances at once, the VMs all have the same name. Click the VM name to rename it appropriately, so that you are able to identify the VMs easily, and press Enter.

  4. Configure the VMs. Select the first instance and click Connect at the top right of the page to connect to the VM console.
  5. Click EC2 serial console and click Connect.
    The VM console appears, the prompt displays the instance ID of the VM.
    Type root as the user, and press Enter.
  6. You are prompted to change your password immediately. Enter and re-enter the new password.
    Note:

    We recommend that you enter the same password for all the VMs. If you configure different passwords for the VMs, enter the different passwords correctly when requested to in step 16 when deploying the cluster.
  7. The console output displays the IP address of the controller VM. Note down the IP address.
    This Controller IP: node1-IP
  8. You are prompted to configure a static IP address.

    Enter n and you are returned to the Deployment Shell with no further IP address configuration needed. Your VM IP address is the same as that assigned by AWS as the controller IP address.

  9. Configure a hostname for the VM.

    1. Type exit to exit to the Linux root shell.

    2. Configure the system hostname using the hostnamectl set-hostname new-hostname1 command.

      Enter an identifying hostname for the VM. For example: primary1. The hostname should be under 64 characters and can include alphanumeric and some special characters.

      The roles of the nodes are assigned automatically. The first three nodes to be configured are the primary and worker nodes. In a four-node cluster, the last node is the worker-only node.

      We do not support changing the hostname after the cluster has been installed.

      When you change the hostname in the Linux root shell, the command prompt changes from the instance ID to the new hostname. However, the command prompt in the Deployment Shell changes only after the cluster deployment is complete.

    3. Type exit to log out of the VM. You are directed to log in to the VM again.

    4. Enter the previously configured password. You are logged in to Deployment Shell.

      Note:

      You are prompted to configure a static IP address every time you log back in to Deployment Shell. Type n and press Enter and repeat until all your nodes are configured and you are ready to deploy the cluster.

  10. Repeat steps 4 through 9 for the other three VMs. Enter appropriate hostnames for the four VMs.
  11. Navigate to EC2 > Instances, click on each Instance ID and note down the private IP addresses of the four node VMs.

You have completed the node preparation steps and are ready to configure the loadbalancer and deploy the cluster. Go to Create the Target Groups and Load Balancers.

Create the Target Groups and Load Balancers

After the VMs are created, configure the target groups and the corresponding load balancers.

Create a Target Group

Targets groups define the target endpoints for the load balancer. Target groups include all cluster VMs and all the defined ports. Each application requires its own target group.

The following table details the target group for each application.
Table 2: Target groups for the load balancer
Target Group Name Protocol Port Health Checks Port for Selected Instances

routing-dir-ui

TCP

30011

TCP

30011

routing-dir-ui-https

TCP

30012

TCP

30012

routing-dir-netconf

TCP

30013

TCP

30013

routing-dir-gnmiterm

TCP

30014

TCP

30014

routing-dir-paa-https

TCP

30016

TCP

30016
routing-dir-paa-https

TCP

30017

TCP

30017

routing-dir-pce

TCP

30018

TCP

30018

routing-dir-rb-crpd

TCP

30019

TCP

30019

routing-dir-rb-ipfix

UDP

30020

TCP (port 30023)

30020

To create a target group for applications detailed in #task_k4w_j3r_y2c__context_bs5_m3r_y2c, perform the following steps:

  1. Navigate to the EC2 > Target Groups page. Click Create new target group at the top right corner of the page.

  2. Select Instances under Choose a target type.

  3. Enter routing-dir-ui as the Target group name.

  4. Select TCP as the Protocol : Port. Enter 30011 as the port number.

  5. Select the VPC on which the cluster is deployed.

  6. Under Health Checks, select TCP as the Health check protocol.

  7. Click Next. The Register Targets page appears.

  8. All the available EC2 instances in your VPC are listed. Select all and only your VMs.

  9. Click Include as pending below. The VMs are registered as the targets and listed under Review targets.

  10. Click Create target group to create the group. The summary of the newly created target group is displayed.

  11. Repeat step 1 through step 10 for all the other target groups listed in Target groups for the load balancer. Enter the appropriate names (3) and port numbers (4) for each target group.

    In step 6, for the routing-dir-rb-ipfix target group, under Health Checks select TCP. Under Health check port, click Override and enter 30023.

Create the Load Balancer

Ensure predictable inbound connectivity by defining traffic paths explicitly through network load balancer listeners and fixed NodePort mappings.
Perform the following steps to create the load balancer, after all the target groups are created.

  1. Click EC2 > Load Balancers. The Load balancers page appears.

  2. To create a load balancer to be used for the GUI, NETCONF, and gNMI-term access, click Create Load balancer at the top right corner of the page. The Compare and select load balancer type page appears.

  3. Click Create under Network Load Balancer.

  4. Enter a name for the load balancer.

  5. Select the VPC on which the VMs are created.

  6. Under Availability Zones and subnets, select the subnet on which the VMs are created. You can also select multiple subnets if your VMs are located in different subnets.

  7. Under Security groups, select the security group that you want to apply on this load balancer. In our example, we select Loadbalancer-sg to apply on the load balancer.

    Also, clear the pre-selected default security group.

  8. Under Listeners and routing, enter the following information.

    1. Select TCP as the Protocol. Enter 80 as the Port number. Select routing-dir-ui as the target group to forward traffic to. For target group names, refer to Target groups for the load balancer.

    2. Click Add listener to add another listener.

    3. Select TCP as the Protocol. Enter 443 as the Port number. Select routing-dir-ui-https as the target group to forward traffic to.

    4. Click Add listener to add another listener.

    5. Select TCP as the Protocol. Enter 2200 as the Port number. Select routing-dir-netconf as the target group to forward traffic to.

    6. Click Add listener to add another listener.

    7. Select TCP as the Protocol. Enter 32767 as the Port number. Select routing-dir-gnmiterm as the target group to forward traffic to.

  9. Click Create load balancer to create the load balancer.

  10. To create a load balancer for the active assurance test agent gateway (TAGW) repeat steps 2 through 7.

  11. Under Listeners and routing, enter following information.

    1. Select TCP as the Protocol. Enter 443 as the Port number. Select routing-dir-tagw-https as the target group to forward traffic to. For target group names, refer to Target groups for the load balancer.

    2. Click Add listener to add another listener.

    3. Select TCP as the Protocol. Enter 6800 as the Port number. Select routing-dir-tagw-https as the target group to forward traffic to.

  12. Click Create load balancer to create the load balancer for the active assurance TAGW.

  13. To create a load balancer for the PCE server repeat steps 2 through 7.

  14. Under Listeners and routing, enter following information.

    1. Select TCP as the Protocol. Enter 4189 as the Port number. Select routing-dir-pce as the target group to forward traffic to. For target group names, refer to Target groups for the load balancer.

  15. Click Create load balancer to create the load balancer for the PCE server.

  16. To create a load balancer for the routing observability repeat steps 2 through 7.

  17. Under Listeners and routing, enter following information.

    1. Select TCP as the Protocol. Enter 17002 as the Port number. Select routing-dir-rb-crpd as the target group to forward traffic to. For target group names, refer to Target groups for the load balancer.

    2. Click Add listener to add another listener.

    3. Select UDP as the Protocol. Enter 4739 as the Port number. Select routing-dir-rb-ipfix as the target group to forward traffic to. For target group names, refer to Target groups for the load balancer.

  18. Click Create load balancer to create the load balancer for the routing observability.

  19. The newly created load balancers are listed on the Load balancers (EC2 > Load balancers) page. Note down the network load balancer URLs listed in the DNS name column. You must enter these hostnames when you deploy the cluster (3).

    You can also create an alias or a custom domain name for a generated DNS name. If you create an alias, you must use the alias during deployment of the cluster. For more information on creating custom domain names, see https://docs.aws.amazon.com/elasticloadbalancing/latest/classic/using-domain-names-with-elb.html.

  20. Resolve the DNS name of the PCE server network load balancer using the nslookup DNS-name-for-PCE-server command on any command line terminal to determine the corresponding IP address.

    Similarly, resolve the DNS names for routing observability CRPD service and IPFIX term.

    You must enter these VIP addresses when you deploy the cluster.

Configure the cluster nodes and deploy the cluster. Go to Deploy the Cluster on AWS.