Backups are an essential part of virtual data protection. Nowadays, thousands of businesses run VMware environments around-the-clock. They all require continuous data protection for such virtualized infrastructures. To deliver efficient backups, let’s review VMware backup practices.

#1 Backing Up Data Regularly

There is a misconception that there is no need to back up virtual machines (VMs). This is because they are not physical machines, but are located on top of the virtualization layer instead. However, this assumption is not entirely true.

Virtual machines and all of the data that resides on them can be affected by ransomware attacks, or system failure. The good news is that modern agile VMware backup solutions can simplify performance and provide instant data recovery options.

#2 Backing Up at the Virtualization Level

Another faulty thought is that guest OS backups are more affordable. Therefore, some businesses implement guest-level instead of host-level backups to save costs. However, host-level backups are now equally affordable, and there is no need to use guest OS backups just to save costs.

Guest-OS backup requires an agent to be installed on the OS of every VM. This agent copies full or partial data during a backup by interacting with the virtualization layer. The host-OS backup, on the other hand, is more streamlined. It can back up data on the block-level at the virtualization layer with no agent required.

#3 Running App-aware Backups

Data consistency is of the highest importance when it comes to backup. Data consistency means that information that you backup is captured in full. That way when you restore it, you have all data in place, meaning that there is no missing or partial information.

App-aware backups ensure that all pending data is included in a backup and new transactions are rolled back to achieve a consistent state. Quiescing is the temporary freezing or stopping of a running VM. This is achieved with the help of Volume Shadow Copy Service (VSS) and VMware tools.

Note: VSS backups are only available for Windows OS. If you use another operating system, such as Linux, you can use the sync driver to quiesce your VMs.  

#4 Snapshots are not Backups

Another mistake is to think that snapshots are similar to backups. If you use a snapshot as a backup, you won’t be able to return your VM to its initial state. Moreover, you will be left with a snapshot from a previous point in time. 

Another issue with snapshots is that they can grow to enormous size if retained for a long time. As new changes occur in the system, the snapshot continues to grow. This puts a huge load on the system and impedes communication between the host and storage area.

#5 Changed Block Tracking (CBT)

Located in the VM kernel, CBT technology is a part of VMware data protection API. With the help of CBT, your backup software can deliver highly efficient backups. CBT ensures fast and space-efficient incremental backups by copying only the blocks that have changed since the last backup.

CBT uses a file (-ctk.vmdk) that stores the mapping of virtual disk blocks in order to track which blocks have changed since the previous backup. CBT improves backup speed, performance and storage capacity and should be implemented by default.

#6 Replicating VMs

A replica is a copy of a production VM. The replica is kept in a secure location referred to as a disaster recovery (DR) site. During a disruption such as a ransomware invasion, natural disaster or system failure, the replica is used to recover production in seconds and avoid downtime. To do this, just go to the DR site and power on the replica.

#7 LAN-Free Data Transfer and Direct SAN

Frequent backups can put a load on your production environment. When a large amount of data travels through the network, the backups may become slower and the overall system may also run less efficiently. The work around for this is to optimize the backup traffic by using LAN-free data transfer, such as Direct SAN access and Hot Add.

  1. Direct SAN. Enables data transfer via Fiber Channel or iSCI from a SAN device. This speeds up and improves the overall efficiency of backup and replication processes.
  2. HotAdd. Enables data reading from VM datastores by using the storage I/O stack. Basically, we “hot add” a snapshot of the source VM as a virtual disk to the VM running the backup app. This bypasses the host’s TCP/IP stack thereby boosting backup speed and performance.

#8 Backup Encryption

Data security should always be kept in mind. Luckily, a modern VMware backup solution can offer extra means for maintaining security. One of those methods is AES-256 data encryption, an international standard for encrypting data. It makes data unreadable during transmission and while in the backup repository.

#9 Backup Automation

Data protection activities take a lot of time. Moreover, backing up data manually often leads to human error. That’s where automation comes in handy. Backup automation is critical for saving time and avoiding sloppy mistakes. Automation also minimizes data access by unauthorized individuals.

A top-notch backup solution helps automate the most vital backup processes, such as scheduling and job chaining. You can also set up to launch an automated recovery after a ransomware attack, disaster or system failure.

#10 Scheduling backups

Running multiple backups at the same time may diminish infrastructure’s performance. System and network overload must be avoided by careful planning. Businesses with large production environments should monitor them and back them up when the workload is minimal.

#11 Keeping an Extra Backup Copy

It’s not enough to just back up a VM. Keep an extra backup copy in a secure secondary location just in case the primary location becomes compromised. Ideally, this repository should be located in a physical location different from your production environment.