Order your Cross-Cloud Interconnect connections

To begin the process of connecting to Amazon Web Services (AWS), order your Cross-Cloud Interconnect connections.

Before you begin

This section describes required permissions, information, and setup steps.

Required roles

Before proceeding, you need the required permissions. Ask your administrator to make sure that you have the Compute Network Admin (roles/compute.networkAdmin) IAM role on the project. For more information about granting roles, see Manage access to projects, folders, and organizations in the Identity and Access Management (IAM) documentation.

Required information

Make sure that you have the names of the remote location and Google Cloud location that you want to use. If you don't have this information, see Choose your locations.

Project selection

If you're using the Google Cloud CLI, set your project ID by using the gcloud config set command.

gcloud config set project PROJECT_ID

The gcloud CLI instructions on this page assume that you have set your project ID.

Capacity planning

When you use Cross-Cloud Interconnect to connect to AWS, you are required to use Link Aggregation Control Protocol (LACP). When you use LACP, you can create an interconnect port "bundle", also known as a link aggregation group (LAG), containing multiple links.

For example, suppose you need 20 Gbps of capacity. In this case, you can order a 20-Gbps port bundle containing two 10-Gbps links. However, it is possible to instead create two separate 10-Gbps bundles, each containing a single 10-Gbps link, if you prefer. Later, when you order your primary and redundant AWS ports, you do so by ordering AWS LAGs. Each LAG represents an AWS DirectConnect port to which you can add multiple links.

One advantage of LACP is that it streamlines the configuration process. For example, as your needs grow, you can add links without having to do as much configuration as you would when adding a new port.

In general, we recommend minimizing the number of Cross-Cloud Interconnect ports and AWS LAGs. Instead, use links to add capacity. However, the best strategy varies depending on your exact capacity needs and the available options. For example, AWS permits up to four links for a 10-Gbps port and up to two links for a 100-Gbps port. So if you need 60 Gbps of capacity, you need to order two 30-Gbps bundle ports with three 10-Gbps links each. You can use equal-cost multipath (ECMP) routing to distribute traffic among them.

For information about LAGs, see Link aggregation groups (LAGs) in the AWS documentation.

Order your connections

When you order a Cross-Cloud Interconnect connection, Google reserves a port on a router at the location that you specify. Google later uses this port to create your connection to an AWS router.

To satisfy the Cross-Cloud Interconnect service level agreement, order two connections: a primary connection and a redundant one.

After you place each order, you get an email confirming your order.

Console

  1. In the Google Cloud console, go to the Cloud Interconnect Physical connections tab.
  2. Go to Physical connections

  3. Click Set up physical connection.
  4. In the form that's displayed, select Cross-Cloud Interconnect connection, and then click Continue.
  5. On the next page, keep the default selection of Order new Cross-Cloud Interconnect connection, and then click Continue.
  6. On the Create Cross-Cloud Interconnect connection page, enter details about the primary connection:
    • Enter a Name for the connection.
    • Optional: Enter a Description.
    • In the Remote cloud provider field, select Amazon Web Services.
    • In the Remote location field, select the AWS location.
    • In the Google Cloud location field, select the Cross-Cloud Interconnect location.
    • Choose a Capacity. Valid values include the following:

      • 10 Gb/s
      • 20 Gb/s (2 x 10 Gb/s)
      • 30 Gb/s (3 x 10 Gb/s)
      • 40 Gb/s (4 x 10 Gb/s)
      • 100 Gb/s
      • 200 Gb/s (2 x 100 Gb/s)

    For help with setting the Remote location or Location field, see Choose your locations.

  7. Click Next.
  8. In the next form, enter details about your redundant connection:
    • Enter a Name for the connection.
    • Optional: Enter a Description.
    • Leave the Remote location and Location fields set to the same locations that you selected for the primary connection.
  9. Click Next.
  10. Enter an email address in the Technical contact field.
  11. Click Next.
  12. Review the order, including the information about billing. If the summary of your details is correct, click Place order. If not, go back and edit the connection details. Then return to this step and place the order.
  13. In the order confirmation dialog, click Submit.

gcloud

Use the gcloud compute interconnects create command.

Cross-Cloud Interconnect is similar to Dedicated Interconnect in the sense that it does not require you to use a partner provider. For that reason, you set the --interconnect-type flag to DEDICATED.

Complete the following steps:

  1. Create the primary connection:

    gcloud compute interconnects create CONNECTION_NAME_1 \
        --interconnect-type=DEDICATED \
        --link-type=CAPACITY \
        --requested-link-count=NUMBER_OF_LINKS \
        --location=GOOGLE_CLOUD_LOCATION_1 \
        --remote-location=REMOTE_LOCATION \
        --description=DESCRIPTION_1 \
        --noc-contact-email=CONTACT_EMAIL_1

    Replace the following:

    • CONNECTION_NAME_1: the name of the primary Cross-Cloud Interconnect connection
    • CAPACITY: the capacity of the connection—for example, LINK_TYPE_ETHERNET_10G_LR or LINK_TYPE_ETHERNET_100G_LR
    • NUMBER_OF_LINKS: the number of Link Aggregation Control Protocol (LACP) links that you want for your connection; valid values are as follows:
      • For 10-Gbps connections: 1 through 4
      • For 100-Gbps connections: 1 or 2
    • GOOGLE_CLOUD_LOCATION_1: one of the two edge availability zones associated with the Google Cloud location you want to use—for example, fra-zone1-58
    • REMOTE_LOCATION: the remote location
    • DESCRIPTION_1: an optional description of your connection
    • CONTACT_EMAIL_1: an email address that Google can use to contact you
  2. Create the redundant connection:

    gcloud compute interconnects create CONNECTION_NAME_2 \
        --interconnect-type=DEDICATED \
        --link-type=CAPACITY \
        --requested-link-count=NUMBER_OF_LINKS \
        --location=GOOGLE_CLOUD_LOCATION_2 \
        --remote-location=REMOTE_LOCATION \
        --description=DESCRIPTION_2 \
        --noc-contact-email=CONTACT_EMAIL_2

    Replace the following:

    • CONNECTION_NAME_2: the name of the redundant Cross-Cloud Interconnect connection
    • CAPACITY: the capacity of the connection; this value must be the same as the capacity of the primary connection
    • NUMBER_OF_LINKS: the number of links that you want for your redundant connection; this value must be the same as the one that you used for the primary connection
    • GOOGLE_CLOUD_LOCATION_2: one of the two edge availability zones associated with the Google Cloud location you want to use; do not use the same zone that you used for the primary location—for example, if you chose fra-zone1-58 for the primary connection, use fra-zone2-58 for the secondary one
    • REMOTE_LOCATION: the same remote location that you specified in the preceding step, when you created the primary connection
    • DESCRIPTION_2: an optional description of your redundant connection
    • CONTACT_EMAIL_2: an email address that Google can use to contact you about the redundant location