Setting up Veeam Replication in a VMware environment

Veeam Backup & Replication is a backup solution and offers a range of features for performing backup tasks and disaster recovery. It is specifically designed for virtual VMware vSphere and Microsoft Hyper-V environments. This article shows how you can use Veeam Backup & Replication to create a replication task in a VMware environment.

Veeam Software at Thomas-Krenn

General information

Veeam Backup & Replication does not require an agent within the VM guest operating system for the replication process as it uses VMware's snapshot functionality to replicate data.^[1]
The VM Snapshot can be viewed as a continuous point-in-time state of a VM, including the configuration of the guest operating system, OS, applications, and files.
Similar to an incremental backup, a complete image of the original VM is transferred to the target host during the first replica cycle.
All subsequent replication cycles are incremental, and Veeam Backup & Replication only transfers data blocks that have changed since the last job was run. Veeam uses Changed Block Tracking technology for this.
With Veeam Backup & Replication you can orchestrate replicas for HA purposes within a site or remote replication across sites for DR purposes.
To enable replication over WAN links or slow connections in general, Veeam uses technologies such as deduplication and compression to optimize traffic.
WAN Accelerators and QOS-like technologies such as Network Throttling also enable fair distribution and efficient use of available resources.

Detailed description of the replication process

The flow of a replication process with Veeam is as described below:

Veeam Backup Manager creates a separate task for each drive to be processed in the replication process.
Veeam Backup Manager connects to Veeam Backup Service, which includes a resource scheduling component that manages all tasks and resources in the backup infrastructure.
Veeam Backup Manager connects to the Veeam Data Mover Services on the backup proxies and backup repository and sets a set of rules for data transfer, such as throttling network traffic, etc.
The source Veeam Data Mover connects to the destination Veeam Data Mover and Veeam Data Mover on the backup repository
Veeam Backup Manager queries information about VMs and ESXi hosts from Veeam Broker Service.
When Application Consistent Processing is enabled, Veeam connects to the guest operating system and runs a "runtime process" to enable Application Consistent Processing.
Veeam Backup & Replication triggers the creation of a VM snapshot through the vCenter Server or ESXi host, putting each VMDK in a read-only state and creating an additional delta file per VMDK. All changes that occur within the VM during replication are written to the delta file.
During replication, the source Veeam Data Mover reads the data from the VM disk and copies it. During a subsequent replication, Veeam uses the CBT information to filter out only the blocks that have changed since the last replication. If CBT information is missing, Veeam uses the information from metadata stored in the repository to detect changed blocks. During the copy process, the source data mover filters out empty blocks, swap files and excluded data to optimize the transfer process. In addition, the data mover compresses the data traffic towards the target data mover.
The data mover on the target side unpacks the data and writes it to the repository.
After the proxy completes the read operation, a snapshot commit prompt is sent to the vCenter or ESXi host to consolidate the snapshot.

Creation of a replication job

The following section shows how to create a replication job:^[2]

Open the New Replication Job Wizard.
In addition to name and description you can make advanced settings on this page. By activating the checkboxes you get access to further setting options in the wizard process about Seeding and Networking.
In this step select the source VMs for replication. Dynamic containers (tags, datastores, pools...) make it easier for you to manage the jobs. These are updated before each replication, for example to recognize new VMs, which are then also replicated.
Select from which source to replicate. In addition to the productive storage, you can also select one or more backup repositories here and thus implement the scenario Remote Replica from Backup, for example.
In this step you can exclude entire VMs or individual VM disks from replication.
To exclude VM discs, select only the discs to be processed based on their disc ID. Unselected discs are not replicated.
Here you define the order in which the VMs are processed by Veeam.
Specify which resources on target side are used for replication, and for example, which disk format should be used for the individual disks of the replica VMs on target side.
If you activated the second checkbox in the first wizard step (Name), you now have the possibility to define settings for Network Mapping.
If you activated the third checkbox in the first wizard step (Name), you now have the possibility to adapt the network schema of your replica VMs to the schema of your DR location using re-IP rules.
Customize the location for replica metadata as well as settings for the name and number of replica points. You will find more customization options under Advanced.
Specify your preferred proxies for this replication job on the source and target side. For traffic-optimized transmission, specify two additional WAN accelerators, one for each side.
If you have activated the first checkbox in the first wizard step (Name) you now have the possibility to benefit from existing backups or replicas on the DR side via Seeding or Mapping.
If you are striving for application-consistent replication of your most important VMs, check the Enable application-aware processing box. Also select a user account with existing rights on the guest operating system.
Use the options provided to set a schedule that makes sense for you and meets your SLA requirements.
In the step Summary the settings you have made are displayed again in summary form and the wizard is completed.