We are constantly questioned by partners and potential clients about the performance of Bacula Enterprise Edition in relation to our competitors, but this is not an easy answer. Distributed backup systems are influenced by numerous infrastructure components, such as:
- Client disk scan speed
- Ethernet traffic bus
- Backup storage write speed
Therefore, we have made this quick comparison to demonstrate that, under equal conditions, Bacula Enterprise is as fast as Veeam, in this case for VMware workloads.
Testing Environment
According to Figures 1 and 2, two virtual machines were used as backup servers on the same Proxmox virtualization host machine. It is possible to observe that the number of processors, cores, RAM, and disk category (NVMe) is exactly the same for both the operating system and the backup target (local storage on the VM itself).
Figures 1 and 2. Windows (Veeam) and Linux (Bacula) Machine Configurations.
According to Figure 3, backups of various VMware machines were performed, with Linux and Windows operating systems:
Figure 3. VMs List of Tests.
Full Backup Comparison (Active Full in Veeam)
According to Figures 4 and 5, the full backup of the same Linux VM (rhel1) takes 01 minute and 19 seconds in Bacula, while the Veeam backup took 02 minutes and 12 seconds. This means that Bacula Enterprise was 33% faster.
Figures 4 and 5. Duration of VMware Linux backups by Bacula and Veeam, respectively.
As shown in Figures 6 and 7, the full backup of the same Windows VM (win2016_2) takes 01 minute and 47 seconds in Bacula, while the Veeam backup took 01 minute and 54 seconds. This means that Bacula Enterprise was approximately 6% faster.
Figures 6 and 7. Duration of VMware Windows backups by Bacula and Veeam, respectively.
Incremental Backup Comparison (Synthetic Full in Veeam)
According to Figures 8 and 9, the incremental backup of the same Linux VM (rhel1) takes 5 seconds in Bacula, while the Veeam backup took 50 seconds. This means that Bacula Enterprise was 1000% faster.
Figures 8 and 9. Duration of VMware Linux incremental backups by Bacula and Veeam, respectively.
As shown in Figures 10 and 11, the incremental backup of the same Windows VM (win2016_2) takes 04 seconds in Bacula, while the Veeam backup took 50 seconds. This means that Bacula Enterprise was 1250% faster.
Figures 10 and 11. Duration of VMware Windows incremental backups by Bacula and Veeam, respectively.
Conclusion
The test results clearly and objectively demonstrate the efficiency and superior performance of Bacula Enterprise Edition compared to Veeam, especially in VMware work environments. A detailed analysis of the figures and data reveals that Bacula is capable of performing full and incremental backups significantly faster, in some cases achieving an advantage of up to 1250% over its competitor. This result is not only a testament to the quality of the software itself but also to the attention given to testing conditions, ensuring equal conditions for a fair comparison. Thus, we can confidently assert that Bacula Enterprise Edition is a solid choice for those seeking high performance and effectiveness in backup and data recovery solutions.
Veeam is a registered trademark of Veeam Software, and the version used in this test was the Community Edition, with original settings.
Disponível em: 
Português (Portuguese (Brazil))
English
Español (Spanish)
