Deploy a Barracuda Backup Virtual Appliance
To create a successful data protection and disaster recovery plan, Barracuda Networks strongly recommends following the 3-2-1 backup rule:
3 – You must have at least 3 copies of your data; the original production data and 2 backup copies.
2 – You must use at least 2 different types of media to store the copies of your data, for example, your production storage and the local Barracuda Backup virtual appliance data storage.
1 – You must keep at least 1 backup offsite, for example, in the cloud or in a remote site.
For deployment of the Barracuda Backup virtual appliance (Vx), Barracuda Networks recommends a completely separate virtual infrastructure from that of the data in which the Vx will be protecting. This means running in a separate virtual cluster, dedicated virtual host, and/or separate storage from the production environment. This will help avoid situations where the production virtual infrastructure or storage fails or is impacted by ransomware and impacts both the production virtual machines and the Barracuda Backup virtual appliance.
The following sections walk you through the process of deploying a Barracuda Backup Virtual (Vx) appliance.
System Platform Requirements
Barracuda Networks provides the following types of virtual images for the Barracuda Backup Vx deployment.
Table 1. System Platform Requirements
Image Type | Platform | Version |
|---|---|---|
OVF | VMware | See Virtual Infrastructure for supported versions. |
VHD | Microsoft Hyper-V | See Virtual Infrastructure for supported versions. |
System Resource Requirements
The most common mistake when designing an environment for Barracuda Backup Vx is to size only for capacity as opposed to sizing for capacity and performance. Tasks such as backup, offsite replication, data restoration, and purging can use a significant amount of system resources. When considering a Barracuda Backup virtual deployment, it is best to consider the use cases of the Backup Vx in addition to needed storage.
Table 2. vCPU Recommendations
Backup Volume | Minimum Recommended vCPUs | High Performance Recommended vCPUs |
|---|---|---|
1-2 TB | 2 vCPUs | 4 vCPUs |
2-6 TB | 4 vCPUs | 6 vCPUs |
6-12 TB | 6 vCPUs | 8 vCPUs |
12-24 TB | 12 vCPUs | 16+ vCPUs |
Table 3. RAM Recommendations
Backup Volume | Minimum Recommended RAM | High Performance Recommended RAM |
|---|---|---|
1-2 TB | 4 GB | 8 GB |
1-4 TB | 8 GB | 12 GB |
4-12 TB | 12 GB | 16 GB |
12-16 TB | 16 GB | 32 GB |
16-24 TB | 32 GB | 64 GB |
Storage Recommendations
Any storage infrastructure consisting of Local/Direct-Attached, SAN, and NAS
Storage with inadequate throughput, read, and write speeds will suffer
Servers with software RAID, slow NAS, and low-RPM drives will yield poor performance
Notes and Considerations
Barracuda Backup utilizes memory (RAM) for database queries during backup, restore, and offsite replication. Barracuda Networks highly recommends following the recommended minimum resources for memory in the RAM Recommendations table above.
Barracuda uses CPU resources for hashing blocks during deduplication, compression for offsite replication, and rehydration of data during restoration. Barracuda Networks highly recommends following the recommended minimum resources for vCPUs in the vCPU Recommendations table above.
At a minimum, memory resources should be dedicated (reserved). It is also recommended that the CPU be dedicated to Barracuda Backup Vx and not shared with other virtual machines on the host. This aligns with VMware and Microsoft’s recommendations for virtualized Microsoft Exchange Server and SQL Server implementations.
A separate data store is recommended for use with the virtual appliance due to disk I/O constraints or an entirely separate storage server/environment altogether from the production environment.
Other system components such as network and storage need to be sized accordingly to prevent them from becoming a bottleneck.
It is important that snapshots are not used on the Barracuda Backup Vx appliance. Even if data is not being replicated to the Barracuda Cloud, configuration data is still sent to Barracuda so that the unit can be managed. Reverting to a previous state using snapshots can potentially cause loss of data and unit corruption. Independent disks are the default selection and highly recommended to avoid snapshots.
Use only a single attached virtual volume when deploying the Barracuda Backup Vx appliance. When expanding Barracuda Backup Vx capacity, expand only the current volume. Adding and expanding a new volume can potentially cause loss of data and unit corruption.
All Backup Vx disk, vCPUs, and RAM can be adjusted later by shutting down the system properly and making the necessary changes through the hypervisor.
VMware vSphere Deployment
To deploy Barracuda Backup Vx in VMware vSphere:
Microsoft Hyper-V Deployment
To deploy a Barracuda Backup Vx in VMware vSphere:
Configure and Provision the Barracuda Backup Vx
You must configure a valid external DNS to resolve to certain Barracuda infrastructure components requiring DNS entries.