Sprint 4C · Capacity planning
Proxmox Backup Storage Calculator
Backup storage is easy to underestimate because retention multiplies changed data while deduplication, pruning and garbage collection work differently from a folder full of full-copy images. This calculator makes the assumptions visible so you can plan repository capacity without pretending PBS savings are a fixed percentage.
Quick answer
How much storage should I allocate to PBS?
Start with the current protected baseline, add expected growth, model how much data changes across the retention schedule, and then apply a conservative dedupe factor only as a planning assumption. Keep additional free space for maintenance, unexpected churn and restores.
Live Amazon hardware
Live repository drives
The catalogue shows current high-capacity HDD and enterprise SSD candidates before the calculator. Capacity, price and seller come from current listings; endurance, PLP and warranty are shown only when the listing actually supports those claims.
Buying decision
Capacity headroom matters more than a perfect dedupe guess
Do not buy a repository that works only if deduplication performs exactly as hoped. A conservative design can absorb a noisy VM, encrypted data, a retention change or a temporary spike in snapshots without immediately becoming a storage emergency.
Interactive planner
PBS Repository Capacity Calculator
Model protected data, retention, change rate, growth and a planning dedupe factor.
The result is an engineering estimate. Real PBS deduplication and compression depend on the workload, encryption, chunk reuse and retention pattern.
Compatibility checkpoints
Check these constraints before buying hardware
Backup independence
Keep the recovery path usable when the primary Proxmox host is unavailable; a separate PBS machine avoids tying restoration to the failed hypervisor.
Repository memory
Size RAM from the datastore as well as the operating system. Treat the documented memory rule as a floor, then leave headroom for concurrent jobs and filesystem cache.
Storage latency
Fast random I/O helps verification, garbage collection and many small chunk operations. HDD repositories benefit from thoughtful metadata and cache architecture.
Network window
Choose the link speed from changed data, concurrency and the time available for backup and restore, not from the nominal size of the server alone.
Start with a measured protected baseline
The capacity model for Proxmox Backup Storage Calculator treats start with a measured protected baseline as a consequence of allocated and actually used guest data rather than marketing capacity. Thin-provisioned virtual disks can make allocated capacity much larger than live data, while databases may change a large share of their blocks every day. Put the numeric assumption beside the retention policy, because storage planning becomes unreliable when change rate or growth is remembered only informally. A repository should remain comfortable even after a noisier workload arrives.
Revisit start with a measured protected baseline after several real backup cycles rather than trusting the first estimate forever. Capture the current used data and its recent growth before choosing a drive set; the calculator is only as useful as the baseline you enter. PBS statistics can replace planning guesses with observed behavior; use the trend to decide whether more disks, faster media or a retention adjustment is actually needed.
Change rate drives retention cost
The capacity model for Proxmox Backup Storage Calculator treats change rate drives retention cost as a consequence of how many gigabytes are genuinely different between recovery points. A 10 TB environment changing one percent per day behaves very differently from a 10 TB database farm rewriting twenty percent. Put the numeric assumption beside the retention policy, because storage planning becomes unreliable when change rate or growth is remembered only informally. A repository should remain comfortable even after a noisier workload arrives.
Revisit change rate drives retention cost after several real backup cycles rather than trusting the first estimate forever. Measure change from real backup statistics when possible and use a higher planning rate for workloads whose churn is still unknown. PBS statistics can replace planning guesses with observed behavior; use the trend to decide whether more disks, faster media or a retention adjustment is actually needed.
Daily, weekly and monthly retention overlap
The capacity model for Proxmox Backup Storage Calculator treats daily, weekly and monthly retention overlap as a consequence of the recovery policy rather than a naive copies-times-size formula. PBS chunk deduplication means retained backups can share unchanged data, so multiplying full capacity by every snapshot overstates requirements. Put the numeric assumption beside the retention policy, because storage planning becomes unreliable when change rate or growth is remembered only informally. A repository should remain comfortable even after a noisier workload arrives.
Revisit daily, weekly and monthly retention overlap after several real backup cycles rather than trusting the first estimate forever. The planning model instead separates a baseline from changed data while still keeping enough margin for imperfect reuse and future churn. PBS statistics can replace planning guesses with observed behavior; use the trend to decide whether more disks, faster media or a retention adjustment is actually needed.
Deduplication is not a guaranteed ratio
The capacity model for Proxmox Backup Storage Calculator treats deduplication is not a guaranteed ratio as a consequence of guest similarity, encryption, compression and data churn. Highly similar VM images can reuse many chunks, while encrypted or already-compressed datasets may share less useful content. Put the numeric assumption beside the retention policy, because storage planning becomes unreliable when change rate or growth is remembered only informally. A repository should remain comfortable even after a noisier workload arrives.
Revisit deduplication is not a guaranteed ratio after several real backup cycles rather than trusting the first estimate forever. Treat a dedupe factor as a sensitivity input and check how the repository plan changes at 1.0x, 1.5x and 2.0x before buying drives. PBS statistics can replace planning guesses with observed behavior; use the trend to decide whether more disks, faster media or a retention adjustment is actually needed.
Growth reserve belongs in the first purchase
The capacity model for Proxmox Backup Storage Calculator treats growth reserve belongs in the first purchase as a consequence of new VMs, larger disks and longer retention. A repository filled to its comfort limit in month one creates pressure to change pruning or add storage immediately. Put the numeric assumption beside the retention policy, because storage planning becomes unreliable when change rate or growth is remembered only informally. A repository should remain comfortable even after a noisier workload arrives.
Revisit growth reserve belongs in the first purchase after several real backup cycles rather than trusting the first estimate forever. Buying headroom at the start can be cheaper than rebuilding the datastore around a second controller or mismatched drive set later. PBS statistics can replace planning guesses with observed behavior; use the trend to decide whether more disks, faster media or a retention adjustment is actually needed.
Pruning and garbage collection are separate steps
The capacity model for Proxmox Backup Storage Calculator treats pruning and garbage collection are separate steps as a consequence of metadata removal versus reclaiming unused chunks. Deleting a backup reference does not instantly mean every associated chunk can be freed because other snapshots may still reference it. Put the numeric assumption beside the retention policy, because storage planning becomes unreliable when change rate or growth is remembered only informally. A repository should remain comfortable even after a noisier workload arrives.
Revisit pruning and garbage collection are separate steps after several real backup cycles rather than trusting the first estimate forever. Leave time and free space for repository maintenance instead of assuming the filesystem shrinks immediately after pruning. PBS statistics can replace planning guesses with observed behavior; use the trend to decide whether more disks, faster media or a retention adjustment is actually needed.
Verification consumes time and I/O
The capacity model for Proxmox Backup Storage Calculator treats verification consumes time and i/o as a consequence of reading chunks to confirm backup integrity. A repository sized only for write ingest can feel slow when verify jobs and new backups compete for the same disks. Put the numeric assumption beside the retention policy, because storage planning becomes unreliable when change rate or growth is remembered only informally. A repository should remain comfortable even after a noisier workload arrives.
Revisit verification consumes time and i/o after several real backup cycles rather than trusting the first estimate forever. Schedule verification realistically and choose SSD or HDD topology based on the maintenance window, not only the initial copy speed. PBS statistics can replace planning guesses with observed behavior; use the trend to decide whether more disks, faster media or a retention adjustment is actually needed.
Restore tests need spare working space
The capacity model for Proxmox Backup Storage Calculator treats restore tests need spare working space as a consequence of temporary target storage and network capacity. A large restore may create another full guest or filesystem copy elsewhere while the repository remains busy. Put the numeric assumption beside the retention policy, because storage planning becomes unreliable when change rate or growth is remembered only informally. A repository should remain comfortable even after a noisier workload arrives.
Revisit restore tests need spare working space after several real backup cycles rather than trusting the first estimate forever. Plan where recovery data lands and ensure the restore destination has enough free capacity before an incident forces that decision. PBS statistics can replace planning guesses with observed behavior; use the trend to decide whether more disks, faster media or a retention adjustment is actually needed.
HDD versus SSD changes cost and maintenance behavior
The capacity model for Proxmox Backup Storage Calculator treats hdd versus ssd changes cost and maintenance behavior as a consequence of capacity economics against random-I/O performance. HDDs reduce cost per terabyte for large archives, while enterprise SSDs can shorten maintenance and concurrent access tasks. Put the numeric assumption beside the retention policy, because storage planning becomes unreliable when change rate or growth is remembered only informally. A repository should remain comfortable even after a noisier workload arrives.
Revisit hdd versus ssd changes cost and maintenance behavior after several real backup cycles rather than trusting the first estimate forever. Use price per usable repository terabyte together with restore time and operational simplicity rather than comparing sticker price alone. PBS statistics can replace planning guesses with observed behavior; use the trend to decide whether more disks, faster media or a retention adjustment is actually needed.
Remote copies change the total storage budget
The capacity model for Proxmox Backup Storage Calculator treats remote copies change the total storage budget as a consequence of second PBS targets, off-site sync and disaster separation. A second repository is intentionally another capacity requirement, but it removes a major site-level failure domain. Put the numeric assumption beside the retention policy, because storage planning becomes unreliable when change rate or growth is remembered only informally. A repository should remain comfortable even after a noisier workload arrives.
Revisit remote copies change the total storage budget after several real backup cycles rather than trusting the first estimate forever. Include remote or offline copies in the infrastructure budget instead of spending every dollar on one enormous local datastore. PBS statistics can replace planning guesses with observed behavior; use the trend to decide whether more disks, faster media or a retention adjustment is actually needed.
Monitor free space as a trend
The capacity model for Proxmox Backup Storage Calculator treats monitor free space as a trend as a consequence of capacity runway rather than a single threshold alert. Knowing that the datastore is 70 percent full is less useful than knowing it has been growing three percent every month. Put the numeric assumption beside the retention policy, because storage planning becomes unreliable when change rate or growth is remembered only informally. A repository should remain comfortable even after a noisier workload arrives.
Revisit monitor free space as a trend after several real backup cycles rather than trusting the first estimate forever. Track repository growth after each retention policy change and re-run the capacity model before the remaining runway becomes short. PBS statistics can replace planning guesses with observed behavior; use the trend to decide whether more disks, faster media or a retention adjustment is actually needed.
Recalculate after workload changes
The capacity model for Proxmox Backup Storage Calculator treats recalculate after workload changes as a consequence of new databases, media libraries or encryption policies. A repository plan can be correct today and wrong after a single high-churn workload is added. Put the numeric assumption beside the retention policy, because storage planning becomes unreliable when change rate or growth is remembered only informally. A repository should remain comfortable even after a noisier workload arrives.
Revisit recalculate after workload changes after several real backup cycles rather than trusting the first estimate forever. Treat the calculator as a repeatable capacity review and keep the assumptions with your backup documentation. PBS statistics can replace planning guesses with observed behavior; use the trend to decide whether more disks, faster media or a retention adjustment is actually needed.
Questions people ask
Proxmox Backup Storage Calculator questions
How accurate is a PBS storage calculator?
It is accurate as a planning model only when the protected-data and change-rate inputs are realistic. Deduplication cannot be predicted as a universal fixed ratio.
Does PBS store a full copy for every backup?
Each backup represents the complete protected state, but PBS can share deduplicated chunks between recovery points and transfers changes efficiently.
What dedupe factor should I enter?
Use 1.0x for a conservative no-savings case, then test modest factors such as 1.5x only if your real PBS statistics support them.
How much free space should a repository keep?
There is no single percentage for every datastore. Keep enough headroom for growth, maintenance, temporary churn and policy mistakes rather than planning to run continuously near full.
Do encrypted guests deduplicate well?
Encryption can reduce cross-backup or cross-guest similarity depending on where encryption is applied and how data changes. Measure rather than assume.
Do snapshots increase PBS storage?
Only changed and newly unique chunks consume additional repository data, but the amount depends on workload churn and retention.
Should I use one huge HDD or several smaller drives?
Choose a redundant filesystem/storage layout that matches performance and failure-recovery needs. One large disk may be simple but creates a single-drive repository failure domain.
Can I mix SSD and HDD backup storage?
Yes in deliberately designed roles, but do not assume an SSD cache automatically solves every HDD bottleneck. Follow the filesystem and PBS storage design you actually deploy.
How do pruning and garbage collection affect space?
Pruning removes backup references according to retention policy; garbage collection later identifies and removes chunks no longer referenced.
Should backup capacity include growth?
Yes. Include expected VM growth and some policy headroom so the repository is not undersized as soon as the environment expands.
Official references and methodology
Verify the current Proxmox and hardware requirements before deployment
The capacity model separates baseline data, changed-data retention, growth and an explicitly user-selected dedupe factor. It does not invent historical savings or present the estimate as a guarantee.
- Proxmox Backup Server system requirements
- Proxmox Backup Server storage documentation
- Proxmox Backup Server FAQ
- Proxmox Backup Server maintenance
As an Amazon Associate, Cloudzat may earn from qualifying purchases. Prices, firmware, transceiver compatibility, link capabilities, switch features, UPS runtime, battery condition and seller terms can change. Verify the exact delivered model and your platform documentation before deployment.