Backup Compression: Protect All Data, Saving Storage and Money
🔎 SUMMARY
– Backup compression is a great way to reduce size of copies.
– Business backups use lossless compression to not lose any detail.
– Storage and network bandwidth savings are the most obvious benefits.
– Not all data types compress easily. And some don’t compress at all.
– LZ4 algorithm offers higher compression speed, while ZStandard one better compression ratio.
Modern organizations generate more and more data. Adding another virtual machine is a matter of a few minutes, hundreds of transactions are saved to your SQL or Oracle database daily in an automated way, and email traffic is ever growing.
Unfortunately, IT budgets are not keeping up the pace, making the purchase of another disk array or cloud storage plan a strain.
There are also the board and CISO who insist on ensuring complete protection, maximum retention, and compliance.
Mission impossible? Well, there’s a way–backup compression. However, you need to remember about the constraints, too.
What is backup data compression?
Data compression is one of the most efficient ways to reduce the size of backup copies. During this process, the file size is reduced through special encoding. To keep it simple, encoding is a smart way of eliminating redundancy and storing information using a shorthand code.
Getting back to compression, there are two main types of compression: lossless and lossy. As the name suggests, lossless compression allows you to shrink the size of data without losing the ability to recover the original data. With lossy compression, the user loses the original data in the process.
Why enterprise backups shouldn’t use lossy compression techniques
Think of lossy compression like summarizing a book—you get the general idea, but some details are gone forever. That’s fine for an unnoticeable movie quality degradation, but fatal for accurate business data. In other words, zero data loss = total accuracy.
Pros and cons of backup compression
Knowing the basics of backup data compression, let’s discuss what benefits it can offer to your organization, and what you should watch out for.
Benefits of compressing backup copies
The most easily measurable effect that the backup compression has on your business is freeing up the storage space. In other words, with compression, you can fit far more data on your existing NAS, SAN, or cloud storage. Keep in mind that, while you can significantly reduce the amount of storage space, the savings may depend on compressed data types. We’ll get back to the topic in a dedicated table below.
And with actual space savings, you also get cost savings. Storage capacity is a cost, no matter if you buy a new SSD drive, a server, or a higher cloud storage plan.
Backup compression helps you save network bandwidth. If you work for a larger organization, your network is possibly very busy. During backup creation, copies travel to backup storage, thus putting an additional load on your infrastructure. By using backup compression, you can greatly decrease the copy size and shorten the time of data transmission, thus offloading your network.
If the two previous aspects (network & storage resources) are your infrastructure’s primary constraints, backup compression can also shorten recovery time and positively impact the Recovery Time Objective (RTO) metric. Bear in mind, however, if computational power, which you need to do data compression maths, is limited, RTO may get even longer.
An indirect benefit of backup compression is a longer retention period. Since compression can—on average—reduce the size of your copies by 30-60% (depending on data types), it can also extend your retention accordingly. Keeping daily or weekly snapshots for a longer time improves your protection against ransomware and historical data loss, allowing you to go back to a specific distant point in time.
Disadvantages of compressing backup copies
Compression is a math-heavy process, which needs significant CPU and memory resources. If you run a backup service/agent on an old infrastructure, compression can make backup tasks run much longer. And if you create backups during business hours, without using the automated scheduler and the backup window feature, compressing data may also impact performance for your end users.
Resources wasted on compressing already compressed data is another disadvantage. If your copy is to include, for example, lots of .zip or .7z archives, .mp4 videos, and .jpeg images, disabling backup compression is the most reasonable step.
What compresses easily and what doesn’t
You know by now that not all data compress well. Study the table below to learn which data types are the best and the worst for compressing. This can help you make the decision whether to enable data compression for a given backup plan/job.
| Type of input data | Example file formats | Compressibility |
|---|---|---|
| Plain text files and source code files | .log, .txt, .csv, .xml, .json | High compressibility |
| Databases | raw SQL dumps (.sql) | |
| Virtual machine images and OS volumes | .vmdk, .vhdx, .vdi, .raw, .iso | |
| Documents (older formats) | .doc, .xls, .ppt | |
| Documents (newer formats) | .docx, .xlsx, .pptx | Moderate compressibility |
| Executables and binaries | .exe, .dll, .bin | |
| Already compressed files (archives) | .zip, .tar.gz, .rar, .7z | Low to zero compressibility |
| Compressed multimedia file types | .mp4, .mkv, .mp3, .jpeg, .png |
Backup compression with Xopero ONE
Xopero ONE, our backup and disaster recovery platform for physical, virtual, and cloud workloads, supports lossless backup data compression out of the box.
The platform allows you to freely customize a backup plan, including data compression options. You can choose between the following two sophisticated compression algorithms:
- LZ4 is a stream compression algorithm, characterized by a very high (near-RAM) speed of compression and even faster decompression. The trade-off is a lower compression ratio compared to the other standard.
- ZStandard is a compression algorithm that works in real-time, providing a high compression ratio. It offers a nice balance between speed and data size. The trade-off? When set for a higher level, it can be resource-consuming.
For an enhanced customization, you can choose between the three compression levels:
- Normal—enables faster backup but is less efficient in terms of saving storage space.
- Medium—the most balanced option. It offers a compromise between the speed and compression ratio.
- High—this one takes longer but reduces the backup size significantly through higher compression ratio.
👉 If you have multiple workloads to protect and want to minimize storage, while maximizing retention, test Xopero ONE and its data compression features for free for 14 days (no credit card required). See how much you can save and gain.
How the backup compression algorithms compare
The table below provides a quick comparison between the LZ4 and ZStandard compression algorithms. In general, the choice depends on whether your infrastructure bottleneck lies within computing power (CPU) or network and storage capacity:
| LZ4 | ZStandard | |
| Compression speed | 🟢 (very high) | 🔴 (high, depending on the level) |
| Decompression speed | 🟢 (very high) | 🟢 (very high) |
| Compression ratio | 🔴 (lower) | 🟢 (higher) |
| Storage consumption | 🔴 (higher) | 🟢 (lower) |
| Cloud/WAN data transmission | 🔴 (typically suboptimal) | 🟢 (usually optimal) |
| CPU usage | 🟢 (very low) | 🔴 (higher and configurable) |
| Backup window focus | minimizing processing time | minimizing data transfer volume |
Best practices for backup compression
Summing up all that’s been said, we’ve prepared a couple of best practices of using backup data compression:
- Assess the storage target and choose:
- ZStandard (Medium or High level) to save costly cloud repository space and optimize internet bandwidth, which is typically the primary bottleneck;
- LZ4 if the storage destination is high-performance local NVMe/SSD disk and the goal is to maximize network throughput.
- Consider computing power (CPU) and choose LZ4 if:
- your environment consists of systems with limited computing resources (CPU, RAM),
- backups are executed during business hours,
- spiking CPU utilization and causing performance degradation for end users must be avoided.
- Consider data change intensity:
LZ4 is recommended when rapid compression is required, and completing the backup quickly to free up the system for subsequent tasks is critical (e.g., tight backup windows or strict Recovery Point Objectives (RPO)). - Consider data fragmentation and choose:
- LZ4 if your backup plan involves a high volume of small files.
- ZStandard if the plan includes larger monolithic files (e.g., VMDK/VHDX virtual machine images, database dumps, ISO files, etc.).
- Opt for ZStandard when planning for:
- long-term data retention,
- backup replication over constrained WAN links.
- Target non-compressible files only
For backup plans with compression enabled, use exclusions to skip file types that are already pre-compressed by default (such as archives and specific multimedia formats). Learn more
