[솔루션] Veeam Backup & Replication (Repliation)

Reference : https://helpcenter.veeam.com/docs/backup/vsphere/replication.html?ver=110

 

 

Replication

 

Replication is a technology that helps you protect mission-critical VMware virtual machines.
When you replicate a VM, Veeam Backup & Replication creates an exact copy of the VM in the native VMware vSphere format on the target host.
Veeam Backup & Replication maintains this copy in sync with the original VM.
Replication provides minimum recovery time objective (RTO) in case a disaster strikes because VM replicas are in a ready-to-start state.

We recommend you to replicate VMs for which recovery point objective (RPO) of hours is required.
If you need RPO of seconds, consider continuous data protection (CDP).

 

Data Replication

To replicate VMs, Veeam Backup & Replication leverages VMware vSphere snapshot capabilities.
During replication, Veeam Backup & Replication requests VMware vSphere to create a VM snapshot.
The VM snapshot is a point-in-time copy of a VM that includes its configuration, OS, applications, associated data, system state and so on. 
Veeam Backup & Replication uses the snapshot as a source of data for replication.

During the first replication cycle, Veeam Backup & Replication copies data of the original VM running on the source host, and creates its full replica on the target host.
Unlike backup files, replica virtual disks are stored decompressed in their native format. All subsequent replication cycles are incremental. 
Veeam Backup & Replication copies only those data blocks that have changed since the last replication job session.
To keep track of changed data blocks, Veeam Backup & Replication uses different approaches. For more information, see Changed Block Tracking.

Veeam Backup & Replication lets you perform onsite replication for high availability scenarios and remote (offsite) replication for disaster recovery scenarios.
To facilitate replication over the WAN or slow connections, Veeam Backup & Replication optimizes traffic transmission.
It filters out unnecessary data blocks such as duplicate data blocks, zero data blocks, blocks of swap files and blocks of excluded VM guest OS files, and compresses replica traffic. 
Veeam Backup & Replication also allows you to use WAN accelerators and apply network throttling rules to prevent replication jobs from consuming the entire network bandwidth.

To replicate a VM, you need to configure required backup infrastructure components and create a replication job.

 

Recovery

If a disaster strikes and the production VM stops working properly, you can fail over to its replica.

When you fail over to a replica, the replica takes over the role of the original VM.
After your original VM is repaired, you can fail back to it and transfer all changes that occurred to replica to the original VM.
If your original VM cannot be repaired, you can perform permanent failover, that is, permanently switch from the original VM to the VM replica and use this replica as the original VM.
For more information, see Failover and Failback for Replication.

 

 

 

 

Backup Infrastructure for Replication

 

Veeam Backup & Replication uses the following components for the replication process:

• Backup server
• Source and target hosts
• Backup proxies
• Backup repository
• [Optional] WAN accelerators

The amount and placement of these components depend on a replication scenario you use. For more information, see Replication Scenarios.

 

Backup Server

During the replication process, the backup server coordinates replication tasks, controls resource allocation and replica job scheduling.
The backup server runs the Veeam Backup Service and Veeam Broker Service that coordinate and interact with the virtual infrastructure.

For more information on the backup server, see Backup Server.

 

Source and Target Hosts

The source host is the host where VMs that you plan to replicate are located.
The target host is the host where VM replicas will be created and maintained in the ready-to-start state.

The role of a target host can be assigned to a standalone ESXi host or ESXi cluster.
If you assign a cluster or vCenter Server as a target, the replication process becomes more sustainable — the replication process will not fail if there is at least one available host in the cluster.

To replicate data from and to hosts, they must be first added to the Veeam Backup & Replication infrastructure.
For more information on how to add hosts, see the Adding VMware vSphere Servers section.

 

Backup Proxies

A backup proxy collects, transforms and transports VM data during the replication process. 
For more information on backup proxies, requirements and limitations for them, see Backup Proxy.

 

For replication, you can deploy backup proxies on the following machines:

• Physical machines.
  In this case, Veeam Backup & Replication uses the Network transport mode to populate replica disk files.
  For more information, see Network Mode.
  
  
• VMs.
  The virtual backup proxy must be registered on an ESXi host that has a direct connection to the target datastore.
  In this case, the backup proxy will be able to use the Virtual appliance transport mode to populate replica disk files.
  This results in increased writing speed and fail-safe replication to ESXi targets.
  For more information, see Virtual Appliance (HotAdd).
  Note that if the Virtual appliance transport mode cannot be used, the backup proxy can fail over to the network mode if you configure it while adding a backup proxy.

 

We recommend you to use at least two backup proxies to ensure that the job will be performed if one of backup proxies fails or loses its connectivity to the source datastore.
For more information on how assign the role of a backup proxy, see Adding VMware Backup Proxies. For more information on how to assign proxies to a replication job, see Specify Data Transfer Settings.

 

Backup Repository

The backup repository stores replica metadata that contains information on the read data blocks (such as checksums and digests).
Metadata is required when Veeam Backup & Replication performs incremental replication or if you fail back from a VM replica to the original VM in the original location using quick rollback. Veeam Backup & Replication uses metadata to quickly detect changed data blocks between two replica states.

The backup repository must be deployed as close to the source backup proxy as possible and must have access to it.

For more information, see Backup Repository.

 

WAN Accelerators

WAN accelerators are optional components in the backup infrastructure. You can use WAN accelerators if you replicate VMs over a slow connection or over WAN.

In the replication process, WAN accelerators are responsible for global data caching and deduplication. To use WAN acceleration, you must deploy two WAN accelerators in the following way:

• The source WAN accelerator must be deployed in the source side, close to the backup proxy.
• The target WAN accelerator must be deployed in the target side, close to the backup proxy.

For more information, see WAN Acceleration.

 

 

How Replication Works

In this article

Veeam Backup & Replication performs VM replication in the following way:

1. When a new replication job session starts, Veeam Backup & Replication reads job settings from the configuration database and creates a list of VMs to process.
   For every disk of a VM added to the job, Veeam Backup & Replication creates a new task.
   
   
2. Veeam Backup & Replication checks what backup infrastructure resources are available, and assigns backup proxies and backup repositories to process the tasks.
   Then Veeam Backup & Replication establishes a connection with source and target backup proxies and the backup repository,
   and sets a number of rules for data transfer, such as network traffic throttling rules and so on.
   
   
3. The source proxy establishes a connection with the target proxy and backup repository.
   
   
4. Veeam Backup & Replication queries information about VMs and virtualization hosts from the vCenter Server.
   
   
5. If application-aware image processing is enabled for the job, Veeam Backup & Replication connects to VM guest OSes, deploys non-persistent runtime components or uses (if necessary, deploys) persistent agent components on VM guest OSes and performs in-guest processing tasks.
   
   
6. Veeam Backup & Replication requests vCenter Server or ESXi host to create a VM snapshot. VM disks are put to the read-only state, and every virtual disk receives a delta file. All changes that the user makes to the VM during replication are written to delta files.
   
   
7. The source backup proxy reads the VM data from the read-only VM disk and copies it. During incremental job sessions, the source proxy uses changed block tracking (CBT) to retrieve only those data blocks that have changed since the previous job session.
   If CBT is not available, the source proxy interacts with the backup repository to obtain replica metadata, and uses this metadata to detect blocks that have changed since the previous job session.

 

While copying VM data, the source proxy performs additional processing.
It filters out zero data blocks, blocks of swap files and blocks of excluded VM guest OS files. The source proxy compresses VM data and transports it to the target proxy.

1. The target proxy decompresses VM data and writes the result to the destination datastore.
   
   
2. After the backup proxy finishes reading VM data, Veeam Backup & Replication requests the vCenter Server or ESXi host to commit the VM snapshot.
   
   

 

 

Requirements and Limitations

In this article

Replication has the following requirements and limitations:

• Due to VMware vSphere limitations, if you change the size of VM disks on the source VM, 
  Veeam Backup & Replication deletes all available restore points (represented as VM snapshots) on the VM replica during the next replication job session.
  For more information, see this VMware KB article.
  
  
• If you assign the role of a backup proxy to a VM, you should not add this VM to the list of processed VMs in a job that uses this backup proxy.
  Such configuration may result in degraded job performance. 
  Veeam Backup & Replication will assign this backup proxy to process other VMs in the job first, and processing of this VM itself will be put on hold. 
  Veeam Backup & Replication will report the following message in the job statistics: VM is a backup proxy, waiting for it to stop processing tasks.
  The job will start processing this VM only after the backup proxy deployed on the VM finishes its tasks.
  
  
• Replication of VM templates is not supported.
  
  
• If you use tags to categorize virtual infrastructure objects, check limitations for VM tags. For more information, see VM Tags.
  
  
• Due to Microsoft limitations, you cannot use Microsoft Azure Active Directory credentials to perform application-aware processing on VMs running Microsoft Windows 10 (or later).
  
  
• If a job is unable to complete within 21 days period, it will be stopped with the Failed status.