Appendix N
Source and Currency Ledger
Every material factual claim on this site has a row here. If something is asserted on a module page, it resolves to one of these entries, and each entry carries a source and the date it was last checked. Claims are not rendered without one — the build fails instead.
The reason this appendix exists first, rather than last, is simple: this guide is written by someone who does not work at Qumulo, about a portfolio that has been changing quickly. The honest way to do that is to show your work, separate what the vendor asserts from what has been independently corroborated, and date all of it.
- Claims tracked
- 36
- Verified
- 19
- Vendor claims
- 12
- Field analysis
- 5
- Estimates
- 0
- Needing re-verification
- 0
How to read a claim type
| Type | What it means | How to use it with a customer |
|---|---|---|
| Verified | Corroborated against a primary source, cited and dated. | Safe to state directly. Still worth re-checking the date. |
| Qumulo states | The vendor asserts it. We report that they assert it. | Attribute it. "Qumulo publishes a figure of…" Never present as measured fact. |
| Field analysis | Independent interpretation. Ours, and labelled as ours. | Offer as a point of view, not as vendor guidance. |
| Estimate | Derived, with assumptions stated in scope. | Show the assumptions or do not use the number. |
The ledger
Public technical depth on Azure Native Qumulo is limited. The Azure Native administrator guide's landing page lists topics for performance characteristics, default limits, and supported configurations, but does not expose provisioning, billing, or scaling specifics at the index level, and this guide publishes no ANQ figures it could not read from a primary source.
ScopeA statement about this guide's sourcing limits, not about the product.
NoteSame discipline as competitor-verification-gaps. Pull ANQ limits and billing detail from the current administrator guide directly before any customer commitment.
Azure Native Qumulo is documented as a Qumulo-managed cloud service on Azure, distinct from the self-managed Cloud Native Qumulo deployments.
ScopeThe managed-versus-self-managed split is the primary decision axis between ANQ and CNQ on Azure, and it drives who owns operations.
The qq CLI is a wrapper that provides a streamlined command-line interface to Qumulo's REST API; Python bindings are available via the qumulo-api package, and the API can also be driven directly with curl.
ScopeEstablishes the API-first management model: anything qq does, the REST API can do, which is the basis for automation conversations.
Qumulo publishes supplementary tools including Power Tools for workload monitoring, a disk-usage aggregation utility, and a parallel copy utility.
ScopeCommunity and vendor tooling; support status varies by tool.
Qumulo reports that its distributed cache serves approximately 95% of all read requests, measured globally across its customer deployments.
ScopeA global fleet-wide aggregate across a customer base skewed toward read-heavy media and research workloads. It is NOT a prediction for any individual customer workload and must not be presented as one. A write-heavy or random-access workload should expect materially different behaviour.
Cloud Native Qumulo is a self-managed deployment of Qumulo Core on AWS, Azure, and GCP, positioned by Qumulo as an exabyte-scale file data platform in the cloud.
ScopeProduct identity and deployment model only. Specific durability, instance, and object-tier design details must be pulled from the current cloud deployment guides, which are versioned and change.
The ratio of data blocks to erasure-coding blocks adjusts as nodes are added, improving storage efficiency at larger node counts while continuing to protect against both disk and node failure. A bit-rot detection algorithm supplements this protection.
ScopeThe vendor does not publish the specific ratios per node count on this page. Do not quote a specific efficiency percentage to a customer without pulling the current supported-configuration documentation.
Qumulo's Global Transaction System makes multi-block writes atomic (all relevant blocks are written, or none are), and the system provides immediate consistency: a write is not acknowledged until data is on persistent storage, so any subsequent read returns a coherent view.
ScopeImmediate (strong) consistency is a meaningful architectural distinction against eventually-consistent object-backed alternatives. It also has performance implications for write-heavy small-file workloads that should be discussed honestly rather than presented as free.
All metadata resides permanently on the storage instance's flash tier.
ScopeRelevant to hybrid configurations, and a direct input to any conversation about small-file and metadata-heavy workload behaviour.
Qumulo tracks aggregated metadata in real time as files and directories are created or modified, propagating changes from individual files up to the root of the file system, which allows analytics queries without a resource-intensive tree walk.
ScopeThis is the mechanism behind Qumulo's real-time capacity and activity analytics. It is the single most load-bearing architectural differentiator to understand before a customer conversation.
Beneath the file system, Qumulo's Scalable Block Store organizes capacity into a protected virtual address space of 4K blocks, each holding either data or erasure-coding hashes.
ScopePhysical clusters. The vendor text scopes this statement to physical deployments.
Qumulo states that a single instance can reach exabytes of capacity and support trillions of files in a unified namespace, without inode depletion or the performance degradation associated with legacy file systems.
ScopeArchitectural ceiling, not a supported-configuration maximum. Never present an architectural ceiling as a supported deployment size.
NoteCAUTION - the vendor page renders this as "exabytes of capacity and 2^64 nodes (~18.4 quintillion files)", which conflates two different units: 2^64 is approximately 18.4 quintillion, but that figure is glossed as "files" while the noun given is "nodes". Do not repeat that sentence as written. Use "trillions of files in a unified namespace", which the same page states separately and which is defensible. Flagged 2026-07-27.
Qumulo states that its containerized architecture allows an entire cluster to be upgraded in 20 seconds, regardless of cluster size.
ScopeRefers to the software swap step. It does not describe end-to-end maintenance-window planning, pre-upgrade checks, or client reconnection behaviour. Verify the current upgrade procedure in the administrator guide before setting a customer expectation.
Qumulo's own public materials describe the Azure offering inconsistently. The Cloud Native Qumulo product page presents Azure as a fully managed SaaS experience integrated into the Azure portal, while the documentation portal separates Cloud Native Qumulo (self-managed on AWS, Azure, and GCP) from Azure Native Qumulo (Qumulo-managed) as distinct offerings.
ScopeObservation about vendor documentation as of this date, not a claim about the products themselves. Both statements may be individually defensible; read together they are confusing.
NoteHIGH FIELD VALUE. Managed versus self-managed is the single most important question in an Azure Qumulo conversation, because it determines who operates the platform, who is on call, and which cost centre pays. Establish which offering is actually on the table before designing anything, and do not resolve the ambiguity from a marketing page.
Cloud Native Qumulo on AWS is documented as a self-managed AWS VPC deployment that provisions EC2 instances and uses AWS infrastructure for a Qumulo file system, a design that allows the disaggregation of persistent storage from compute resources. Deployment is a two-step sequence: create persistent storage using S3 buckets, then deploy cluster compute and cache resources.
ScopeAWS specifically. The two-step deployment order is the architecture made visible: capacity exists independently of, and outlives, the compute that serves it.
Qumulo documents operational limits for Cloud Native Qumulo on AWS, including a maximum cluster size of 24 nodes when deployed via CloudFormation, a requirement that gp3 volumes with custom IOPS and throughput settings complete EBS optimization before running high-performance workloads, and the fact that S3 buckets are subject to throttling prior to partition scale-up.
ScopeThese are the numbers that matter in a design review and they are the ones least likely to appear on a slide. The S3 partition-scale-up throttling behaviour in particular is a real early-life performance characteristic: a benchmark run immediately after deployment may not represent steady state.
NoteRe-verify the 24-node CloudFormation limit before sizing. Documented limits change between releases far faster than architecture does.
Qumulo states that Cloud Native Qumulo separates compute from capacity using a mirrored NVMe cache tier for fast writes paired with cloud object storage, with a caching layer that anticipates data access to deliver NVMe-class latency.
ScopeVendor description of the caching architecture. "NVMe-class latency" is a characterisation, not a measured figure, and applies to cache hits.
Qumulo's documentation states that Cloud Native Qumulo on AWS provides the same multi-protocol support, interfaces, and functionality as Qumulo on premises, including NFS, SMB, and S3 API access.
ScopeFunctional parity of protocols and interfaces. It is NOT a statement about performance parity, which depends entirely on instance selection, network, and object-tier behaviour.
Qumulo states that cloud-native instances can scale throughput beyond 100 GBps and capacity to hundreds of petabytes.
ScopeCeiling claim. Achieved throughput is a function of instance type, instance count, network, object-tier behaviour, and workload shape. Any customer number requires a sized configuration and, realistically, a POC.
Pure Storage, Inc. changed its corporate name to Everpure, Inc., filing a certificate of amendment with the Delaware Secretary of State on 20 February 2026, effective 23 February 2026. The company's stock ticker changed from PSTG to P on the New York Stock Exchange, with trading under the new symbol beginning 17 April 2026.
ScopeCorporate identity only. Product names in the FlashBlade family were not renamed by this change. Corroborated independently by Forbes, Computer Weekly, and Diginomica coverage, and directly observed: purestorage.com URLs now 301-redirect to everpuredata.com.
NoteHIGH FIELD VALUE. Any competitive matrix, battlecard, or slide written before late February 2026 names this competitor incorrectly. Saying "Pure Storage" in a 2026 customer meeting signals that your competitive material is more than six months stale, which is a credibility cost out of all proportion to the error.
Everpure positions FlashBlade as a scale-out blade-and-chassis platform supporting NFS, SMB, and S3 on a single system, in three lines: FlashBlade//E for cost-optimised active archive and backup, FlashBlade//S for data protection, AI training and inference, analytics and HPC at up to 650 GB/s, and FlashBlade//EXA for extreme-scale AI and HPC at 10+ TB/s.
ScopeVendor-published positioning and headline throughput figures, read directly from the vendor page. Ceiling numbers for specific configurations, not workload-independent guarantees. Do not compare these to a Qumulo figure without matching the configuration, the workload, and the measurement method - that comparison is meaningless otherwise.
Dell PowerScale OneFS is a scale-out NAS operating system that stripes data across all nodes in a cluster and supports NFS, SMB, HTTP, FTP, S3, and HDFS. A cluster comprises a minimum of three F-series nodes. Dell's April 2026 white paper describes nodes carrying up to 736 GB of RAM each, giving a cluster up to 181 TB of system memory at 252 nodes.
ScopeRead directly from the H8202.27 PDF (April 2026 revision). IMPORTANT PRECISION - the 252 figure appears in a sentence about aggregate cache memory, NOT as a stated maximum supported cluster size. Do not quote "252 nodes maximum" as Dell's supported ceiling on the strength of this document; it supports "a cluster of 252 nodes is described" and nothing stronger.
NoteOneFS is the closest architectural analogue to Qumulo Core and the most common incumbent you will displace. Treat it with respect in front of a customer who already owns it - they chose it, and disparaging it disparages them.
VAST Data's public web pages describe the DASE architecture in positioning terms - a clean-slate, disaggregated shared-everything design - but route architectural specifics to gated video, analyst reports, and solution-engineer conversations rather than publishing them on the open web.
ScopeAn observation about public information availability as of this date, not a criticism of the product, and not evidence of any weakness in it.
NotePractical consequence - you cannot responsibly build a detailed VAST comparison from public sources alone. If a deal is genuinely against VAST, get current technical detail through a legitimate channel rather than inferring it. Inferred competitive detail is how an SA gets publicly corrected in front of a customer.
Dell's and NetApp's marketing sites returned HTTP 403 to automated retrieval during this source pass, so NetApp AFX node counts, capacity ceilings, and AI Data Engine specifics are NOT independently verified in this guide and no figures for them are published here.
ScopeA statement about the limits of this guide's sourcing, not about the products. NetApp's disaggregated AFX direction is real and well covered in trade press, but this guide does not publish specific AFX numbers it could not read from a primary source.
NotePull these figures from the vendor's current documentation directly before any customer use. A competitive claim you cannot source is a claim you cannot defend when the incumbent's SE is in the next meeting.
WEKA's WekaFS is a distributed parallel file system that bypasses the kernel for lower latency, composed of frontend client services, backend compute handling data distribution, protection, metadata and tiering, and an SSD drive agent. It supports POSIX, NFS, SMB, S3, and GPUDirect Storage, and tiers cold data to object storage while keeping hot data and metadata on NVMe.
ScopeRead directly from WEKA's official documentation. GPUDirect Storage support is the axis on which WEKA most often wins GPU-training evaluations; if that is the customer's requirement, say so rather than contesting it.
WEKA announced NeuralMesh 6 on 21 July 2026, describing it as the most significant software release in the company's history, with general availability in the second half of 2026. NeuralMesh is now the name of WEKA's overall platform. Announced capabilities include multi-tenancy up to 50,000 logically isolated tenants per cluster, unified S3 and POSIX on NVMe with zero-copy GPU transfer, and always-on data reduction claimed at up to 6x capacity savings.
ScopeANNOUNCED, NOT GENERALLY AVAILABLE as of this verification date. Comparing a shipping Qumulo release against a competitor's announced-but-unreleased capability is exactly the kind of unfair comparison this guide forbids (PRD S18). If a customer raises NeuralMesh 6, establish whether they are evaluating what ships today or what is promised for H2 2026.
NoteNaming collision worth being ready for - WEKA's platform is "NeuralMesh" while Qumulo ships NeuralCache, NeuralProtect and NeuralSearch. A customer hearing both in the same evaluation may conflate them. Be the person who disambiguates clearly rather than the person who exploits the confusion.
The unstructured-data competitive landscape changed materially in the first seven months of 2026: Pure Storage became Everpure in February, NetApp moved toward a disaggregated ONTAP architecture with AFX, and WEKA consolidated on NeuralMesh as its platform identity with a major release announced in July.
ScopeSynthesis of the individually sourced claims in this section. Offered as field analysis, not vendor guidance.
NoteThe practical lesson for a field guide reader is not the individual facts but the rate of change - competitive material has a shelf life measured in months, and the failure mode is not being wrong about architecture, it is being confidently out of date about names.
On its Cloud Native Qumulo page, Qumulo states NeuralCache achieves cache hit rates over 96%, up to 100x faster time-to-first-byte compared to native object storage, and can cut transaction fees by over 99% through write compaction before writes reach object storage. The same page states support for workloads requiring 1 TBps or more of sustained throughput.
ScopeAll four are vendor figures with no published methodology, workload definition, or measurement date. The 100x time-to-first-byte comparison is against raw object storage, which is a low bar by design - that is the entire reason a cache tier exists. The transaction-fee reduction is the most commercially interesting of the four because API request charges are a real and frequently underestimated line item in cloud file architectures, but it still requires the customer's own access pattern to mean anything.
NoteNOTE THE TENSION with arch-cache-hit-rate-95, which cites ~95% from the technical architecture page as a global fleet figure. This page says "over 96%" for NeuralCache. Different scopes, different pages, similar-sounding numbers. Do not blend them into one claim, and do not present either as a prediction for a specific customer.
Qumulo Nexus documentation is listed on the documentation portal with restricted access.
ScopeMeans public technical depth on Nexus is limited. Treat any detailed public Nexus capability description with caution and attribute it.
Qumulo's published portfolio comprises the platform architectures Qumulo Core ("distributed data operating system"), Cloud Native Qumulo, Qumulo Stratus ("data platform for zero-trust, AI-native environments"), and Qumulo Nexus ("AI-assisted management portal"); plus the platform features Cloud Data Fabric, NeuralCache ("AI-enabled predictive caching and prefetch engine"), NeuralProtect ("real-time ransomware and malware protection"), NeuralSearch ("storage-native search for unstructured data at scale"), and Qumulo Accelerators ("lightweight data engines for remote and constrained environments").
ScopeNames and vendor taglines as published on 2026-07-27. Product naming in this portfolio has moved quickly; re-verify before any customer-facing use.
NoteNote the vendor's own split between "platform architectures" (things you deploy) and "platform features" (capabilities within them). Getting this hierarchy right is most of what separates someone who understands the portfolio from someone reciting a feature list.
Qumulo states that in Stratus each tenant has dedicated encryption keys and data is encrypted before reaching shared storage, so multi-tenant isolation is cryptographically enforced rather than administratively enforced.
ScopeThe distinction between cryptographic and administrative isolation is architecturally meaningful and worth explaining to a security buyer. It is also exactly the kind of claim a security team will want evidence for - key management, key custody, and rotation are the follow-up questions, and this guide does not have public answers to them.
Qumulo describes Stratus as a disaggregated, shared-nothing design with two independent layers: Stratus Accelerators, stateless high-performance compute nodes that run the Qumulo file system and handle NFS, SMB, S3 and encryption; and Stratus DataCore, a shared durable storage foundation optimised for capacity and cost-efficiency.
ScopeVendor-published architecture description. The word that matters is "stateless" - it is what allows the compute layer to be scaled, replaced, or lost without implicating the durable layer, and it is the same separation CNQ achieves in the cloud, brought on premises.
Qumulo states it has shipped 7 exabytes of capacity and serves customers in 56 countries.
ScopeCompany-published cumulative figures as of 2026-07-27.
Qumulo states it serves more than 1,100 enterprise customers globally.
ScopeCompany-published figure as of 2026-07-27. Qumulo is a private company and does not publish audited customer counts.
Qumulo states a Net Promoter Score of 95.
ScopeSelf-reported, with no published methodology, sample size, or measurement date. An unusually high NPS for any enterprise infrastructure vendor. Report it as the company's claim; do not present it as independently validated.