Azure Standard HDD: Performance and Cost Guide Standard HDD is still the cheapest managed disk tier in Azure. For dev/test boxes, backup targets, and rarely touched data, it often does the job. But the lowest price per GiB doesn't guarantee the lowest total cost.

Slow boots, throttled backups, and a pager going off at 2 a.m. all land on the bill. So do orphaned disks, extra snapshots, and tier rounding. Flexera's 2026 State of the Cloud estimates 29% of cloud spend is wasted.

This guide covers what Standard HDD can do, how to test your workload, how to model the full cost, how it compares with SSD tiers, and how to keep disk spend under control.

Key Takeaways

  • Standard HDD suits low-cost, latency-tolerant workloads, not production databases.
  • Published limits are "up to" figures; validate them against current Microsoft docs.
  • Billing covers provisioned tier, transactions, snapshots, and retained disks.
  • HDD OS disks retire September 8, 2028; data disks are unaffected.
  • Ongoing visibility and rightsizing prevent disk waste.

Azure Standard HDD Explained

Standard HDD is a managed block disk backed by hard disk drives. Microsoft positions it for latency-tolerant workloads such as backup, dev/test, and noncritical, infrequently accessed data. It is not recommended for low-latency OS or transactional database workloads.

Tiers and limits

All figures are "up to" values, per disk. Larger tiers can reach expanded targets with performance plus.

Tier GiB Baseline IOPS Baseline MB/s Expanded IOPS / MB/s
S4–S20 32–512 500 60 None listed
S30 1,024 500 60 1,500 / 150
S40 2,048 500 60 3,000 / 300
S50 4,096 500 60 3,000 / 500
S60 8,192 1,300 300 3,000 / 500
S70–S80 16,384–32,767 2,000 500 3,000 / 500

Azure Standard HDD tiers showing capacity and performance limits

Note that S4 through S50 share the same baseline ceiling. Buying a bigger disk in that range adds capacity, not speed.

Redundancy, lifecycle, and fit

  • HDD disks are locally redundant (LRS) only. There is no ZRS option.
  • Standard HDD OS disks retire September 8, 2028, with automatic conversion to Standard SSD of equal size. Migrate beforehand to avoid disruption.
  • No retirement has been announced for HDD data disks.
  • Unmanaged disks are a separate matter and were retired March 31, 2026.

Good fits include dev/test, lightly used apps, low-I/O servers, and cost-driven bulk data. It isn't a substitute for Blob or Archive storage, which serve different access models.

Azure Standard HDD Performance: What to Expect

Metrics that matter

  • IOPS and throughput: your disk's published IOPS and MB/s limits. Hitting them shows up as slow backups and delayed jobs.
  • Average and tail latency: tail spikes cause sluggish app responses and slow boots.
  • Queue depth: a growing queue means the disk can't keep up.
  • Burst: disk bursting is available only on Standard SSD and Premium SSD. Standard HDD uses Performance Plus as a separate feature.

The VM can be the real limit

Effective performance is the lower of the disk limit and the VM limit. A Standard_D2s_v5 caps uncached remote storage at 3,750 IOPS and 85 MB/s baseline, while a D8s_v5 allows 12,800 IOPS and 290 MB/s. Microsoft's own example shows three disks requesting 15,000 IOPS but a VM limiting delivery to 12,800.

HDD behavior also varies more than SSD with I/O size, sequential versus random access, read/write mix, and concurrency. Don't expect SSD-like consistency.

Test with a representative VM

  1. Deploy a VM of the same size you'll use in production.
  2. Attach the target HDD tier.
  3. Run DiskSpd (Windows) or fio (Linux) with your real block sizes, read/write ratio, and concurrency.
  4. Run long enough to capture variability, not just a quick peak.
  5. Compare against production observations.

Five-step Azure Standard HDD workload testing process

Microsoft's published benchmark examples use Premium SSD, so none of them are Standard HDD guarantees. Treat your own test as one data point, not a promise.

Monitor after deployment

Track these Azure Monitor metrics:

  • Data Disk IOPS and Bandwidth Consumed Percentage
  • Queue Depth and Latency
  • VM-level uncached consumed metrics

A reading of 100% means you've reached the disk or VM ceiling. Watch over a representative period and set thresholds from your own application response times, not generic numbers.

Azure Standard HDD Cost: Pricing, Billing, and Total Cost of Ownership

How billing works

Azure bills Standard HDD by provisioned tier, not by filesystem usage. A 1 TiB disk is billed as S30; grow it to 3 TiB and billing moves to S50. A disk just over a boundary pays for the next tier.

Transactions are charged per 10,000 billable transactions, covering reads, writes, and deletes. The billing guide describes 16-KiB transaction units, but the pricing page applies that only in select regions. Check your region on the official Managed Disks pricing page and the Azure pricing calculator. Rates change often, so confirm current figures on the pricing page and calculator.

Build the full monthly model

Keep these lines separate:

  • Disk tier charge
  • Transaction charges
  • VM compute
  • Snapshots and backup
  • Data transfer
  • Monitoring
  • Retained or unattached disks

Azure managed disk monthly cost components and billing model

Snapshot billing differs by type. Full snapshots bill on used data. Incremental snapshots start with a full copy, then store deltas, and are billed on used size.

Compare by workload, not per GiB

A cheaper disk that slows a batch job, triggers throttling, or needs engineering time to work around can cost more overall. Factor in:

  • Performance needs
  • Transaction volume
  • Availability expectations
  • Engineering effort
  • Cost of a slow or failed workload

Common cost traps

  • Oversized disks provisioned "just in case"
  • Disks left behind after VM deletion, which keep billing
  • Unattached, unmounted, or zero-I/O volumes
  • Snapshots kept long past their usefulness
  • Redundancy choices nobody reviewed

Microsoft's FinOps guidance recommends removing or downgrading unattached disks. Azure Advisor flags them too, though deletion is irreversible, so snapshot first.

Standard HDD vs Standard SSD, Premium SSD, and Ultra Disk

All maximums are per disk.

Type Best for Max IOPS Max MB/s OS disk Redundancy
Standard HDD Backup, noncritical 2,000 (3,000 with Performance Plus) 500 Yes, until 2028 retirement LRS
Standard SSD Web, light enterprise apps, dev/test 6,000 750 Yes LRS/ZRS
Premium SSD Low-latency production 20,000 900 Yes LRS/ZRS
Premium SSD v2 Sustained high-performance production 80,000 2,000 No LRS
Ultra Disk I/O-intensive databases 400,000 10,000 No LRS

Azure managed disk types performance and redundancy comparison

Premium SSD v2 and Ultra are available only in selected regions, so verify before designing around them.

When Standard SSD wins

Choose Standard SSD when the app needs steadier latency but not Premium IOPS. Microsoft meets Standard SSD performance targets 99% of the time, versus 99.9% for Premium SSD.

That consistency gap, plus ZRS support, is often worth the modest price difference for web servers and lightly used business apps. It is also the natural landing spot for HDD OS disks before 2028.

When to go Premium or Ultra

Production databases, transaction-heavy apps, AI workloads, and strict performance requirements belong on Premium SSD, Premium SSD v2, or Ultra. Remember that v2 and Ultra are data-disk-only, so pair them with a Premium SSD OS disk.

How to Decide and Optimize Azure Disk Spend

Start with an inventory

For every disk, record:

  • Purpose and business criticality
  • Used capacity and growth trends
  • I/O pattern and latency needs
  • Backup needs
  • Whether it is attached, mounted, and generating I/O

A simple decision framework

  • Standard HDD: genuinely low-I/O, cost-sensitive workloads
  • Standard SSD: general-purpose needs with more consistent performance
  • Premium SSD: production and performance-sensitive systems
  • Ultra Disk: only when its performance and compatibility needs are real

Size for capacity and headroom

Check VM-level limits, growth forecasts, and recovery objectives.

Resizing has real constraints:

  • Online expansion works for eligible data disks
  • Crossing 4 TiB on smaller disks requires deallocation
  • Shrinking is unsupported

Switching HDD to SSD means stopping the VM, changing the account type, and restarting, so plan the downtime. Because you can't shrink, overprovisioning is hard to undo.

Govern continuously

  • Tag disks and assign owners (resource-group tags do not inherit automatically; use Azure Policy)
  • Review utilization on a schedule
  • Clean up idle disks
  • Set snapshot retention rules
  • Validate backups
  • Set budget alerts
  • Retest workloads before and after any tier change

Where Lucidity fits

Manual reviews struggle at scale. Lucidity offers Lumen, a storage observability tool for enterprise teams across Azure and multicloud environments. It surfaces unattached, reserved, unmounted, and zero-I/O disks, which Lucidity's internal data shows can account for up to 70% of unused block storage spend. In one 372-disk example, disks tied to terminated instances cost $71.3K a year and never-attached disks $37.4K.

Lucidity Lumen identifying unused Azure block storage across environments

Lumen bases tier recommendations on historical IOPS, throughput, latency, and cost trends rather than snapshots in time. Lucidity's AutoScaler expands disks as data grows with no downtime. It complements Azure's tiers rather than replacing them.

Conclusion: Is Azure Standard HDD Right for You?

Azure Standard HDD makes sense when access is infrequent, latency needs are modest, and the workload tolerates variable performance. It is a poor match for latency-sensitive or I/O-heavy production systems.

Before you commit:

  1. Confirm current Microsoft limits and lifecycle status.
  2. Model the full monthly cost.
  3. Benchmark a representative workload.
  4. Verify VM-level constraints.
  5. Monitor performance after deployment.
  6. Review disk utilization regularly to spot idle capacity.

Frequently Asked Questions

How much does Azure Disk Storage cost?

It depends on disk type, provisioned tier, region, redundancy, transactions where applicable, and related services like snapshots. Use the Azure Managed Disks pricing page and pricing calculator for current rates.

What are the key differences between Azure Standard SSD and Premium SSD?

Standard SSD fits web servers and lightly used apps, with up to 6,000 IOPS and 750 MB/s per disk. Premium SSD targets low-latency production, with up to 20,000 IOPS and 900 MB/s, and a tighter 99.9% performance target.

Which Azure disk type cannot be used as an OS disk?

Premium SSD v2 and Ultra Disk are data-disk-only. Standard HDD, Standard SSD, and Premium SSD can serve as OS disks, though HDD OS disks retire in 2028. Check restrictions for your VM and region.

Is Azure Standard HDD suitable for production workloads?

Only for lightly used, latency-tolerant systems. Databases and user-facing applications generally need SSD tiers. Match the choice to I/O, latency, availability, and recovery requirements.

Can I change an Azure Standard HDD to another disk type?

Yes. For Standard SSD or Premium SSD, deallocate the VM, change the account type under Size + performance, and restart. Conversion to Premium SSD v2 or Ultra follows separate, restricted paths. Back up first and follow Microsoft's latest procedure.