2.12 Regular maintenance of the components
A critical environment designed to offer a service to external customers like Veeam Cloud Connect needs to guarantee the best possible experience and the highest possible uptime. For this reason, different operational criteria should be applied when managing Veeam Cloud Connect.
As outlined in this chapter, Veeam Cloud Connect is designed as a distributed architecture. To guarantee the best performance, each component can and should be deployed on a separated server, physical or virtual.
Even if multiple components can be installed in the same server, when a maintenance activity is needed or an issue is found, the downtime created can negatively affect each component deployed in the same server. If the Veeam Backup & Replication server is deployed together with one of the cloud gateways, the Veeam server will need stopped when maintenance is needed for a gateway, which creates downtime for the entire server.
By splitting each component over a different server and deploying several instances of those components that allow multiple instances, the chances to interrupt the entire Veeam Cloud Connect service because of a single unavailable server are extremely reduced.
Patches and upgrades
Veeam Cloud Connect is deployed for the vast majority of its components over Microsoft Windows operating systems. In order to guarantee stability and protection from vulnerabilities (especially for those components exposed to internet like the cloud gateways), service providers should check regularly for the availability of Microsoft patches and upgrades, and apply them as soon as possible. Again, the distributed architecture allows for a stacked patching activity of different components at different times.
Service providers also need to address the Veeam Backup & Replication updates. If the Veeam Backup & Replication server is able to reach Veeam Update Notification server (http://dev.veeam.com), the software will notify users about the availability of a new update.
Service providers that subscribe to receive email notifications from the VCSP program should also receive notifications about new updates a few weeks in advance when compared to the notification server. This notification system enables service providers to plan upgrades to their environment before each user sees the available update from the notification server.
Veeam Cloud Connect is backwards-compatible to previous versions up to one major release. For example, service providers using v9 can receive backups from customers using v8. However, a service provider cannot run a version older than what the customer uses regardless of this compatibility. The service provider should deploy any update before it’s generally available to end users.
Time sync and DNS resolution
Two important aspects of every network are time sync and DNS resolution. In a Veeam Cloud Connect environment, these two services must be properly configured and monitored to guarantee correct operations.
Every server of the environment — both those running Veeam Cloud Connect components and the additional machines like Active Directory servers or the hypervisor hosts — must be synchronized with the same time source, and providers must be sure that time and time zone are correctly configured. Differences in time between the different components can lead to unexpected errors when operating Veeam Cloud Connect.
The same is true for DNS: Each and every component of the environment must be reachable by using forward DNS resolution. The DNS servers must be correctly configured and populated with every existing record (for both forward and reverse zones), and they must check to verify that they are operating correctly and are reachable over the network by any other server. Redundant DNS servers should be deployed to guarantee optimal uptime of the overall DNS service. If you follow this guide, you’ll see two Windows Servers as both Active Directory controllers and DNS servers for the entire infrastructure.