What we do.

This matters because storage decisions at research-computing scale are rarely about a single product. An institution standing up a multi-petabyte storage platform is simultaneously making decisions about the file system, the tape archive, the hardware fleet, the operational practice, the procurement strategy, and the long-term cost trajectory — and those decisions interact. A file-system choice that ignores the archive tier produces a system that cannot retain data affordably. A hardware-procurement strategy that ignores the operational practice produces a fleet that cannot be sustained. Unique Checksum exists to make these decisions coherently, which is why the offerings are unified rather than siloed.

The sections that follow describe each pillar in turn — what it delivers, how it connects to the others, and where the deeper detail lives on the rest of this site. The flagship pillar is the most developed, because LStore is the platform around which the firm is currently organized; the tape and consulting pillars complete the picture of what an institution can engage Unique Checksum to do.

This is the role from which the rest of the flagship offering follows: Unique Checksum does not merely resell access to a file system; it coordinates the documentation, roadmap, hardware specification, and operational-practice work that turns a research-and-engineering project into a platform other institutions can deploy and sustain. The architectural authorship of LStore belongs to Vanderbilt and ACCRE; the work of making it deployable beyond Vanderbilt is what Unique Checksum leads.

What an institution engages under this pillar:

• Platform access Solution-set access to LStore for institutions outside Vanderbilt — the distributed file system, its tooling, and the deployment knowledge required to run it in production.
• Deployment & operations design The placement policy, reliability configuration, lifecycle orchestration, and depot operations practice that match the platform to the institution’s workload — the work the LStore sub-pages describe in technical detail.
• Hardware specification Guidance on the hardware that runs LStore well. Unique Checksum maintains the certified-compatible-hardware list for LStore, drawing on the production fleet’s demonstrated tolerance for heterogeneous, multi-generation hardware.
• Partner ecosystem Access to the broader partner ecosystem forming around LStore. The platform is the center of a solution set that extends beyond the file system itself, and Unique Checksum is the coordination point for it.

The “& partners” in the pillar name is deliberate. LStore is not a single-vendor product; it is the center of a solution ecosystem, and the partners who extend it — in orchestration, in hardware, in adjacent capabilities — are part of what an institution gets when it engages the platform through Unique Checksum. That ecosystem is actively forming, and the pillar is named to hold room for it.

The tape pillar exists because the storage lifecycle does not end at disk. Research-computing institutions accumulate data with retention horizons measured in decades — regulatory obligations, research-reproducibility requirements, the long tail of data that cannot be deleted but does not justify spinning disk. For that data, modern tape is the right medium: the highest capacity per cartridge, the lowest cost per terabyte, the longest archival shelf life, and a genuine air gap that no online system can provide. Spectra Logic builds the tape library platforms Unique Checksum offers for this purpose.

Unique Checksum scopes this pillar to the tape libraries specifically — the physical preservation platform — and integrates it with LStore as the archive tier in the storage lifecycle. When an institution deploys LStore with a Spectra Logic tape target, the result is a single bundled solution: the distributed file system handling active data across its tiers, and the tape library handling long-term preservation, with LStore staging data to tape when placement policy determines it belongs there. The integration is architectural, described on the Operations & Lifecycle page, and made operational by Unique Checksum’s own tape-library-management intellectual property.

The full treatment of the Spectra Logic partnership — the tape library family, the role tape plays in the LStore solution, and the preservation case — is on the dedicated Spectra Logic page.

The consulting practice draws on the same heritage as the rest of the firm. Twenty-five years inside high-performance and research-computing storage — building integrity technology, running it in production, and watching which institutional practices succeed at scale and which do not — is the basis for the advice. The practice is not generic IT consulting; it is specific to the problems that large-data research computing actually faces, brought by people who have operated these systems rather than only studied them.

• Storage architecture advisory Guidance on the architecture of multi-petabyte and exabyte-scale storage — tiering strategy, file-system selection, reliability design, and the cost trajectory of the choices, independent of which platform an institution ultimately deploys.
• Operations practice design The operational discipline that sustains storage at scale: the depot operations cadence, the integrity-verification practice, the failure-response procedures that turn a deployed system into a reliable one.
• Vendor & partnership management Facilitation of the vendor and partnership relationships that large storage deployments require — including federal partnership facilitation for institutions operating in the government-adjacent research space.
• Procurement strategy Strategy for acquiring storage and infrastructure economically — the hardware-mix decisions, the refresh-cycle planning, the procurement-vehicle navigation that determine whether a system is affordable over its lifetime.

The consulting practice is also where engagements often begin. An institution that is not yet ready to deploy a platform may engage Unique Checksum first for the architecture and strategy work, and arrive at a platform decision through that engagement rather than before it. The advisory pillar is the front door as often as it is a standalone service.

The shape of an engagement follows the problem. A research-computing center standing up its first large-scale distributed storage platform engages differently from a government IT organization adding a tape archive to an existing system, which engages differently again from a commercial enterprise seeking architecture advice before committing to a direction. Unique Checksum scopes each engagement to its actual objectives, with deliverables and acceptance criteria defined up front so that both parties know what success looks like.

For LStore platform engagements specifically, the structure of the relationship — how Unique Checksum works with an institution’s internal storage team, the operational cadence, the division of responsibility between Unique Checksum and the institution — is described in detail on the Engagement Models page. That page is the right next read for an institution seriously evaluating a platform deployment.

The common thread across the three sectors is not the type of institution but the nature of the problem: data large enough, valuable enough, and long-lived enough that the decisions about where it lives and how its integrity is assured carry real consequence. Unique Checksum’s offerings are organized around that problem, which is why they apply across sectors that otherwise look quite different from one another.

This is why the offerings are unified rather than siloed. The LStore platform expresses the discipline architecturally — per-file integrity verification, erasure-coded durability, continuous operational practice. The tape pillar extends it to the preservation horizon. The consulting practice brings it to institutions that have the data but not yet the practice. And the same intellectual-property estate that underlies the firm — the Constancy backup system in production at ACCRE, the tape-library-management technology extended into LStore — reflects the same philosophy at its engineering core. Different surfaces, one discipline, one owner.

For institutions deciding whether to engage, the question Unique Checksum would pose is the one that runs through the whole site: not which product to buy, but whether the institution’s data is being kept with the discipline its scale and value require. The three pillars are three ways of answering that question affirmatively. The path into any of them begins with a conversation.