Block Storage

Name: Block Storage
Brand: Google
Rating: 4.3 (42 reviews)

Block Storage is a portfolio of network-attached and locally-attached block devices for Google Cloud Compute Engine VMs and Google Kubernetes Engine, providing durable Persistent Disk volumes, tunable Hyperdisk volumes, and ephemeral Local SSDs for different performance and durability needs.

Est 1998|Valuation $4081.3B|Raised $26.1M

LinkedIn 41.9M|Website|Web traffic 3.3M|FTEs 198,933|HQ Mountain View, California, US

MarketsCloud Infrastructure Providers (IaaS/PaaS)Cloud Infrastructure Providers (IaaS/PaaS)

Published Oct 12, 2026

01Common complaints

Company sizes or stages where this product is overkill or insufficient
- Very small teams may incur avoidable costs if they provision persistent disks and forget to delete/unattach them; pay-as-you-go models can produce ongoing charges for idle resources. (https://www.reddit.com/r/googlecloud/comments/1redt46/3_gcp_quick_wins_that_saved_us_2kmo_took_under_an/)
Industries or use cases where competitors are clearly more appropriate
- If an organization is standardized on AWS or Azure and lacks a strategic reason to run workloads on GCP, using EBS or Azure Managed Disks may reduce organizational friction.
Technical requirements that this product cannot meet
- If the newest GCE machine families are required and Persistent Disk is unsupported on them, PD is not an option; Hyperdisk or an architecture change is necessary. (https://cloud.google.com/compute/docs/disks/)
- If temporary ultra-low-latency scratch storage is required and data loss on stop/restart is acceptable, Local SSD may be more appropriate than durable block storage. (https://cloud.google.com/compute/docs/disks/)
Budget constraints that make alternatives more appropriate
- If data can be stored as objects (Cloud Storage) or moved to managed databases, overall VM + disk footprint may be reduced, although this changes architecture and operational model.
Specific workflows or integrations where the product falls short
- Multi-node read-write shared disk patterns can be complex; even when multi-writer exists, many applications require a clustered filesystem or coordination to avoid corruption.
Red flags in the buying process
- If clear answers on disk availability by machine series/region and performance ceilings for a chosen VM type are not obtainable, there is a risk of re-architecture during deployment.
Current limitations acknowledged by the vendor
- Persistent Disk availability constraints
  - Google states Persistent Disk is not available with the latest machine series and recommends Hyperdisk where possible. (https://cloud.google.com/compute/docs/disks/)
- Network latency nature
  - Persistent Disk is networked storage and generally has higher latency than physical disks/Local SSD.
- Attach limits and multi-attach restrictions
  - Documentation notes per-VM disk attachment limits and constraints (e.g., read-only multi-attach limits for certain PD types). (https://docs.cloud.google.com/compute/docs/disks/persistent-disks)
Features on the roadmap (not yet available)
- No explicit roadmap dates were found in the sources accessed.
Scale limitations (at what volume/size does it break down?)
- PD per-disk size limit cited as up to 64 TiB; per-VM total disk limits depend on machine type (documentation notes up to 128 PD volumes for many VMs, with total size constraints). (https://docs.cloud.google.com/compute/docs/disks/persistent-disks)
Security or compliance gaps
- Encryption is documented; explicit certification mapping at the product-page level was not found in retrieved sources.
Technical debt or architectural limitations mentioned in reviews
- Reviews/community discussion point more to operational tuning and observability gaps than architectural technical debt.
API limitations or integration constraints
- Not deeply assessed in retrieved sources; buyers should validate API/CSI feature parity for their chosen disk type (PD vs Hyperdisk) and region.

- Company sizes or stages where this product is overkill or insufficient - Very small teams may incur avoidable costs if they provision persistent disks and forget to delete/unattach them; pay-as-you-go models can produce ongoing charges for idle resources. (https://www.reddit.com/r/googlecloud/comments/1redt46/3_gcp_quick_wins_that_saved_us_2kmo_took_under_an/) - Industries or use cases where competitors are clearly more appropriate - If an organization is standardized on AWS or Azure and lacks a strategic reason to run workloads on GCP, using EBS or Azure Managed Disks may reduce organizational friction. - Technical requirements that this product cannot meet - If the newest GCE machine families are required and Persistent Disk is unsupported on them, PD is not an option; Hyperdisk or an architecture change is necessary. (https://cloud.google.com/compute/docs/disks/) - If temporary ultra-low-latency scratch storage is required and data loss on stop/restart is acceptable, Local SSD may be more appropriate than durable block storage. (https://cloud.google.com/compute/docs/disks/) - Budget constraints that make alternatives more appropriate - If data can be stored as objects (Cloud Storage) or moved to managed databases, overall VM + disk footprint may be reduced, although this changes architecture and operational model. - Specific workflows or integrations where the product falls short - Multi-node read-write shared disk patterns can be complex; even when multi-writer exists, many applications require a clustered filesystem or coordination to avoid corruption. - Red flags in the buying process - If clear answers on disk availability by machine series/region and performance ceilings for a chosen VM type are not obtainable, there is a risk of re-architecture during deployment. - Current limitations acknowledged by the vendor - Persistent Disk availability constraints - Google states Persistent Disk is not available with the latest machine series and recommends Hyperdisk where possible. (https://cloud.google.com/compute/docs/disks/) - Network latency nature - Persistent Disk is networked storage and generally has higher latency than physical disks/Local SSD. - Attach limits and multi-attach restrictions - Documentation notes per-VM disk attachment limits and constraints (e.g., read-only multi-attach limits for certain PD types). (https://docs.cloud.google.com/compute/docs/disks/persistent-disks) - Features on the roadmap (not yet available) - No explicit roadmap dates were found in the sources accessed. - Scale limitations (at what volume/size does it break down?) - PD per-disk size limit cited as up to 64 TiB; per-VM total disk limits depend on machine type (documentation notes up to 128 PD volumes for many VMs, with total size constraints). (https://docs.cloud.google.com/compute/docs/disks/persistent-disks) - Security or compliance gaps - Encryption is documented; explicit certification mapping at the product-page level was not found in retrieved sources. - Technical debt or architectural limitations mentioned in reviews - Reviews/community discussion point more to operational tuning and observability gaps than architectural technical debt. - API limitations or integration constraints - Not deeply assessed in retrieved sources; buyers should validate API/CSI feature parity for their chosen disk type (PD vs Hyperdisk) and region.

Deal disqualifiers: what makes a POOR fit

Requirement for ultra-low latency akin to local NVMe without application-level caching (Local SSD or specialized architectures may be needed)
Requirement to run on the newest machine series where Persistent Disk is unavailable (alternative storage approaches or a redesign may be required; see official vendor documentation)

Why some customers left for competitors (churn reasons)

Direct "we left for X" narratives were not prominent in the accessible sources, but themes that can drive churn include:
- Price sensitivity versus commodity alternatives
- Need for higher maximum performance or independent performance tuning (leading to internal alternatives or cross-cloud moves)
- Operational friction with Kubernetes storage

02Why they buy

Competitive advantages validated by customer feedback (not just vendor claims)
- Ease of attaching and maintaining disks with GCE VMs is frequently mentioned in reviews (review excerpts note straightforward disk attachment to VMs). (https://www.g2.com/products/google-cloud-persistent-disk/reviews)
Unique features and capabilities competitors lack
- Hyperdisk Storage Pools and thin-provisioning/data reduction are described by Google as potential management and TCO levers; these are vendor statements and should be validated for specific workloads. (https://cloud.google.com/products/block-storage)
Differentiation from competitors based on reviews
- Review data in accessible sources is limited; the clearest differentiator is Google’s product direction: Hyperdisk for tunable performance and higher ceilings, PD for compatibility.
Value propositions with evidence from case studies
- Limited hard evidence in accessible sources; most quantified claims in sources retrieved are vendor statements.
Areas where competitors clearly have advantages (be balanced)
- AWS and Azure may offer broader third-party tooling integrations and organizational familiarity, especially in enterprises already standardized on those clouds.

- Competitive advantages validated by customer feedback (not just vendor claims) - Ease of attaching and maintaining disks with GCE VMs is frequently mentioned in reviews (review excerpts note straightforward disk attachment to VMs). (https://www.g2.com/products/google-cloud-persistent-disk/reviews) - Unique features and capabilities competitors lack - Hyperdisk Storage Pools and thin-provisioning/data reduction are described by Google as potential management and TCO levers; these are vendor statements and should be validated for specific workloads. (https://cloud.google.com/products/block-storage) - Differentiation from competitors based on reviews - Review data in accessible sources is limited; the clearest differentiator is Google’s product direction: Hyperdisk for tunable performance and higher ceilings, PD for compatibility. - Value propositions with evidence from case studies - Limited hard evidence in accessible sources; most quantified claims in sources retrieved are vendor statements. - Areas where competitors clearly have advantages (be balanced) - AWS and Azure may offer broader third-party tooling integrations and organizational familiarity, especially in enterprises already standardized on those clouds.

Why customers chose this product over alternatives (as suggested by sources)

Tight integration with Google’s compute ecosystem (GCE + GKE), ease of attaching disks to VMs, and generally positive perceptions of reliability. (Official vendor documentation and product overview pages)

03Field reports

Review platforms (G2)

Overall sentiment on G2 appears positive but based on a very small sample size, which limits statistical confidence.
Examples of positive/neutral feedback (verbatim excerpts)
- G2 reviewer (Mar 29, 2022): "Simple and reliable storage" and "It does what it says." (G2 product review page)
- G2 reviewer (Jul 26, 2022): "Easy to maintain and integrate" (G2 product review page)
Examples of criticisms (verbatim excerpts)
- G2 reviewer (Mar 29, 2022): "Pricing is not the most attractive…" (G2 product review page)
- G2 reviewer (Jul 26, 2022): "Need more stat/reports…" (G2 product review page)
Notable limitation
- The accessible G2 set did not surface lower-star reviews in the snippet (distribution showed predominantly high-star ratings), which makes it harder to balance with negative review-platform feedback.

- Overall sentiment on G2 appears positive but based on a very small sample size, which limits statistical confidence. - Examples of positive/neutral feedback (verbatim excerpts) - G2 reviewer (Mar 29, 2022): "Simple and reliable storage" and "It does what it says." (G2 product review page) - G2 reviewer (Jul 26, 2022): "Easy to maintain and integrate" (G2 product review page) - Examples of criticisms (verbatim excerpts) - G2 reviewer (Mar 29, 2022): "Pricing is not the most attractive…" (G2 product review page) - G2 reviewer (Jul 26, 2022): "Need more stat/reports…" (G2 product review page) - Notable limitation - The accessible G2 set did not surface lower-star reviews in the snippet (distribution showed predominantly high-star ratings), which makes it harder to balance with negative review-platform feedback.

Review platforms (Gartner Peer Insights)

Access and summary
- A Gartner Peer Insights listing for "Persistent Disk" is accessible with an overview and rating context (favorable average rating with a moderate number of ratings shown on the page excerpt; service information recently updated). Full individual review text may require login/subscription and may be partially inaccessible.
Example excerpt available on listing
- The listing describes Persistent Disk as block storage for VM instances and highlights data encryption at rest, automated backups, and scalability. (Gartner product listing)

- Access and summary - A Gartner Peer Insights listing for "Persistent Disk" is accessible with an overview and rating context (favorable average rating with a moderate number of ratings shown on the page excerpt; service information recently updated). Full individual review text may require login/subscription and may be partially inaccessible. - Example excerpt available on listing - The listing describes Persistent Disk as block storage for VM instances and highlights data encryption at rest, automated backups, and scalability. (Gartner product listing)

Community/Social (Reddit)

Performance frustrations (standard PD)
- A user reports a large I/O test taking much longer on a mid-sized VM with a modest standard persistent disk versus their laptop, implying a mismatch between expectations and PD performance scaling with size/type. (Community Reddit post)
Reliability/anomaly reports
- A thread reports an SSD persistent disk failure on a Compute Engine instance, with symptoms like intermittent access gaps and the disk effectively becoming unavailable at times (user-reported; not an incident postmortem). (Community Reddit post)
GKE attach/mount operational issues
- A thread shows Kubernetes events with FailedMount/timeouts and a "cannot mount unformatted disk…" scenario (likely configuration/workflow error, but it reflects operational friction). (Community Reddit post)
Cost hygiene
- A community post on saving costs explicitly calls out unattached persistent disks as a cost win. (Community Reddit post)

- Performance frustrations (standard PD) - A user reports a large I/O test taking much longer on a mid-sized VM with a modest standard persistent disk versus their laptop, implying a mismatch between expectations and PD performance scaling with size/type. (Community Reddit post) - Reliability/anomaly reports - A thread reports an SSD persistent disk failure on a Compute Engine instance, with symptoms like intermittent access gaps and the disk effectively becoming unavailable at times (user-reported; not an incident postmortem). (Community Reddit post) - GKE attach/mount operational issues - A thread shows Kubernetes events with FailedMount/timeouts and a "cannot mount unformatted disk…" scenario (likely configuration/workflow error, but it reflects operational friction). (Community Reddit post) - Cost hygiene - A community post on saving costs explicitly calls out unattached persistent disks as a cost win. (Community Reddit post)

Specific outcomes customers achieved (with metrics when available)

Vendor-stated metrics
- The vendor page claims Hyperdisk can deliver very high throughput and IOPS from a single Hyperdisk and that Hyperdisk Storage Pools can reduce storage TCO by a significant margin. These are vendor statements; methodology and workload assumptions are not shown and should be validated via independent benchmarks and pricing models for a specific configuration. (Official vendor product and feature pages)

Case Studies

Vendor-provided case study signals (limited)
- The Persistent Disk product page displays customer logos (e.g., Etsy appears in the page snippet) but the excerpt accessed does not provide a full, independently verifiable case study with metrics. (Official vendor product overview and feature pages)
Data protection feature case study (product launch)
- The Archive Snapshots launch announcement (Sept 12, 2022) frames the feature as a lower-cost approach for long-term retention; it is a product capability announcement, not an audited customer ROI study.
- Critical evaluation
  - Pricing benefits depend on restore frequency, retention requirements, and compliance workflows.
  - The post describes retrieval charges and billing-period constraints; buyers should model those carefully. (Official vendor blog and product launch announcement)
Independent/press context for product evolution
- Independent press coverage from the period (e.g., mid-2014 reporting) covered Google adding SSD persistent disks; historically it reported no extra I/O request fees at the time and highlighted performance differences per storage unit. This is historical context; current pricing/performance should be validated against current docs. (Independent press coverage and archival reporting)

- Vendor-provided case study signals (limited) - The Persistent Disk product page displays customer logos (e.g., Etsy appears in the page snippet) but the excerpt accessed does not provide a full, independently verifiable case study with metrics. (Official vendor product overview and feature pages) - Data protection feature case study (product launch) - The Archive Snapshots launch announcement (Sept 12, 2022) frames the feature as a lower-cost approach for long-term retention; it is a product capability announcement, not an audited customer ROI study. - Critical evaluation - Pricing benefits depend on restore frequency, retention requirements, and compliance workflows. - The post describes retrieval charges and billing-period constraints; buyers should model those carefully. (Official vendor blog and product launch announcement) - Independent/press context for product evolution - Independent press coverage from the period (e.g., mid-2014 reporting) covered Google adding SSD persistent disks; historically it reported no extra I/O request fees at the time and highlighted performance differences per storage unit. This is historical context; current pricing/performance should be validated against current docs. (Independent press coverage and archival reporting)

04Best fit

Company sizes

Mid-market to large enterprise organizations that:
- Run production workloads on Google Cloud (GCE/GKE)
- Need durable storage attached to compute (databases, stateful services)

Industries and verticals where adoption is strongest

SaaS, technology, media, fintech—any vertical already adopting GCE/GKE
Regulated industries can use it, but should validate compliance scope at the broader Google Cloud platform level

Roles and personas

Decision makers
- Head of Infrastructure / Platform Engineering
- Cloud Center of Excellence
- Security/Compliance (for encryption key management and auditability)
End users
- SREs, DevOps engineers
- Kubernetes platform operators
- Database administrators

Pain points and challenges the product addresses

Durable boot/data disks independent from VM lifecycle (detach/attach, resize)
High availability options (regional replication across multiple zones)
Snapshot-based backup patterns

Qualification criteria: what makes a GREAT fit

Workloads already on GCE or GKE
Need straightforward, cloud-native durable block devices
Willing to benchmark and tune performance (queue depth, disk type selection)

05The pitch

What the product does (core value proposition in objective terms)

Google Cloud “Block storage” is a portfolio of network-attached and locally-attached block devices used primarily with Compute Engine VMs and Google Kubernetes Engine (GKE). Google describes three main options:
- Hyperdisk (a newer network-attached block storage option)
- Persistent Disk (PD; a legacy network-attached block storage option)
- Local SSD (ephemeral, locally-attached NVMe/SSD) — see product overview and feature pages.
Persistent Disk volumes behave like disks attached to a server but are accessed over the network; data is distributed across multiple physical disks for redundancy and performance management. See official product documentation for implementation details.

- Google Cloud “Block storage” is a portfolio of network-attached and locally-attached block devices used primarily with Compute Engine VMs and Google Kubernetes Engine (GKE). Google describes three main options: - Hyperdisk (a newer network-attached block storage option) - Persistent Disk (PD; a legacy network-attached block storage option) - Local SSD (ephemeral, locally-attached NVMe/SSD) — see product overview and feature pages. - Persistent Disk volumes behave like disks attached to a server but are accessed over the network; data is distributed across multiple physical disks for redundancy and performance management. See official product documentation for implementation details.

Target market and market positioning

Target users
- Infrastructure / platform teams running VMs and stateful containers on GKE.
- Database and enterprise application operators who need durable volumes (boot + data) with predictable performance.
Positioning
- Google promotes Hyperdisk as the recommended durable block storage option where it is supported, with Persistent Disk serving legacy/compatibility needs on older VM families or where Hyperdisk is not available. See vendor product documentation for current guidance.

- Target users - Infrastructure / platform teams running VMs and stateful containers on GKE. - Database and enterprise application operators who need durable volumes (boot + data) with predictable performance. - Positioning - Google promotes Hyperdisk as the recommended durable block storage option where it is supported, with Persistent Disk serving legacy/compatibility needs on older VM families or where Hyperdisk is not available. See vendor product documentation for current guidance.

Get a comprehensive analysis of Block Storage including market position, competitive landscape, adoption trends, and peer benchmarks.

Download full report →