Migrating Classic Service Points to Fleet Service Points - BlueCat DNS Edge

DNS Edge User Guide (Fleet Service Point)

Locale
English
Product name
BlueCat DNS Edge
BlueCat recommends migrating all Classic Service Points to the Fleet Service Point to take advantage of the modern platform and features that will be introduced in future releases. The following sections outline methods for switching DNS clients from Classic Service Points to Fleet Service Points.
The following configurations are currently not supported with Fleet Service Points:
  • DSR load balancing due to lack of support for IP aliases on Fleet Service Points.
  • Modifying the private IP address of the Fleet Service Point on AWS and GCP due to limitations with the Cloud provider.

Anycast migration

If you are running the Classic Service Point in an Anycast environment, you can provision a new Fleet Service Point and configure Anycast service on the Fleet Service Point. Once that has been configured, you can add the Fleet Service Point to the Anycast pool and deploy the DNS resolver service to use the Anycast IP address of the Fleet Service Point. Once you have provisioned the required Fleet Service Points that will replace the Classic Service Points, test them to ensure that they are configured correctly. You can then decommission the Classic Service Points in the Edge Cloud UI and delete the Classic Service Point VMs. With this approach, DNS clients would not experience any downtime.

For more information on the procedures outlined in this section, refer to the following documents:

IP address swap migration on vCenter and ESXi

For Classic Service Points that are configured on vCenter and ESXi, and are not part of an Anycast environment, you can perform an IP address swap where the provisioned Fleet Service Point takes over the IP address of the Classic Service Point. To perform the IP address swap, provision the Fleet Service Point using a temporary IP address and deploy the DNS resolver service to the Fleet Service Point using the same site configuration as the Classic Service Point. Once you have tested the Fleet Service Point to ensure that it is configured correctly, you can then decommission the Classic Service Point in the Edge Cloud UI and delete the Classic Service Point VM. Once you have removed the Classic Service Point, reconfigure the IP address on the Fleet Service Point to match that of the Classic Service Point that was deleted.

For more information on the procedures outlined in this section, refer to the following documents:
Once you have decommissioned the Classic Service Point and deleted the Classic Service Point VM, perform the following steps to reconfigure the IP address on the Fleet Service Point:
  1. Log in to the Fleet Service Point using a console session.
    Attention: BlueCat recommends connecting to the Fleet Service Point using a console session, as the change in network configuration can result in the Fleet Service Point becoming unreachable using an SSH connection.
  2. Modify the IP address, netmask, and default gateway of the eth0 interface within /etc/network/interfaces.d/50-cloud-init to match the interface settings of the Classic Service Point.
    Note: A warning appears at the top of the file stating that the contents of the file will be overwritten upon reboot. This message can be safely ignored, as the VMware network configuration only runs once for each Fleet Service Point and not upon each reboot.
  3. Save the changes to the file.
  4. Run the following command to reconfigure the eth0 interface with the new settings and restart the Fleet platform services:
    fleet newIP

    All services will restart and listen on the new IP address of the Fleet Service Point.

While the Fleet platform services restart, the Fleet Service Point will be in an Unhealthy state for a few minutes. During this time, DNS clients will experience a brief downtime.

IP address swap migration on Azure

For Classic Service Points that are configured on Azure, you can perform an IP address swap where the provisioned Fleet Service Point takes over the IP address of the Classic Service Point. To perform the IP address swap, provision the Fleet Service Point using a temporary IP address and deploy the DNS resolver service to the Fleet Service Point using the same site configuration as the Classic Service Point. Once you have tested the Fleet Service Point to ensure that it is configured correctly, you can then decommission the Classic Service Point in the Edge Cloud UI and delete the Classic Service Point VM. Once you have removed the Classic Service Point, reconfigure the IP address on the Fleet Service Point to match that of the Classic Service Point that was deleted.

For more information on the procedures outlined in this section, refer to the following documents:
Once you have decommissioned the Classic Service Point and deleted the Classic Service Point VM, perform the following steps to reconfigure the IP address on the Fleet Service Point:
  1. Log in to the Azure portal.
  2. Click Virtual machines.
  3. Click the name of the Fleet Service Point virtual machine that you are reconfiguring the IP address on.
  4. Click Networking in the side bar and click the name of the network interface attached to the VM.
  5. Click IP configurations in the side bar and click the private IP address of the VM.
  6. Change the IP address of the VM to match the interface settings of the Classic Service Point.
  7. Click Save.

    Once you have clicked saved, the VM reboots. After the VM has completed rebooting and its core services have restarted, all services will restart and listen on the new IP address of the Fleet Service Point.

The Fleet Service Point takes at least 5 minutes to reboot on Azure and services take an additional two minutes to restart after the reboot has completed. While the Fleet platform services restart, the Fleet Service Point will be in an Unhealthy state for a few minutes. During this time, DNS clients will experience a brief downtime.

Note: During the restart process, you might see similar messages to the following within the serial console of the VM followed by no logging output for several minutes:
SMBus base address unitialized - upgrade BIOS or use force_addr=0xaddr
This is expected behaviour and does not indicate that an error has occurred during the reboot process.

Migrating Classic Service Points on AWS and GCP

Due to limitations with the Cloud provider, you cannot modify the private IP address of the Fleet Service Point on AWS and GCP. If you have Classic Service Points deployed to AWS or GCP, you must first decommission the Classic Service Point in the Edge Cloud UI and delete the Classic Service Point VM. Once you have removed the Classic Service Point, provision a new Fleet Service Point VM on your AWS or GCP environment with the IP address reclaimed from the Classic Service Point, and deploy DNS resolver service to the Fleet Service Point using the same site configuration as the Classic Service Point. With this approach, DNS clients would experience a brief downtime during the Fleet Service Point provisioning and DNS resolver service deployment process.

For more information on the procedures outlined in this section, refer to the following documents: