Is the File System Dying?

CHESA Fest 2026 brought together technology vendors, media organizations, and workflow architects to explore the architectural shifts reshaping modern content infrastructure. As part of the event, a series of vendor panels examined the deeper technical debates emerging across storage, asset management, and AI-driven workflows.

This discussion focused on one of those foundational debates; how the rise of object storage and cloud-native architectures is challenging long-standing assumptions about the role of the file system in media production, and whether traditional file-based workflows remain the
operational backbone of modern environments or are evolving into a performance layer within an increasingly object-native ecosystem.

Is the File System Dying?

The Performance Tier in an Object-Native World

For years, the file system has been the unquestioned center of gravity in media production. If you were editing, finishing, transcoding, or archiving, you mounted a volume and got to work. It wasn’t debated. It
was assumed.

But that assumption is quietly being tested.

Object storage now underpins nearly every cloud workflow. SaaS creative tools are training a generation of professionals to think in applications, not directories. APIs are becoming first-class citizens inside production software. And at CHESA Fest 2026, Vendor Panel 1 took the question head-on:

If applications can increasingly interact directly with object storage, what happens to the file system?

Not whether object storage works.

Not whether cloud workflows are viable.

But whether the file system remains the architectural core, or becomes something more specialized inside a broader object-native stack.

The Panel

The discussion was moderated by Tom Kehn, Vice President of Solutions Consulting at CHESA, and brought together storage and workflow leaders from across the media technology ecosystem to examine how modern production environments are evolving as object storage becomes more deeply integrated into creative workflows.
Panelists included:

• Rich Werhun, Senior Solutions Engineer at LucidLink
• Ryan Servant, Senior Director of Channel & Alliances at Suite Studios
• Dave Simon, Senior Director of Media & Entertainment Alliances at Backblaze
• Nathan Halverson, Manager of Solutions Architecture at Spectra Logic

The conversation also included contributions from additional practitioners and audience members, including Dave Helmly, Director of Strategic Development for Professional Video at Adobe, whose
perspective from the application layer helped frame how creative tools may evolve as storage architectures shift.

Together, the group explored a question quietly emerging across the industry: as object storage continues to scale and applications increasingly interact with data through APIs rather than mounted volumes, does the traditional file system remain the center of gravity for media production workflows, or is it evolving into a performance layer within a broader object-native architecture?

The discussion quickly moved beyond simple “file versus object” comparisons and into deeper territory: how modern workflows balance performance, governance, lifecycle management, and the expectations
of creative users who increasingly care less about where data lives and more about how quickly they can access it.

The File System as Abstraction

The first meaningful pivot in the discussion came when Rich Werhun of LucidLink reframed the premise entirely.

“I see the file system as the abstraction layer. That’s what it is.”

That statement reshapes the debate.

If object storage continues to win on scalability and economics—and it clearly is—something still has to translate object semantics into something creative tools can consume. Even if applications eventually speak S3 natively, they won’t be the only systems interacting with that data.

Workflows are ecosystems. They include transcoders, QC tools, AI engines, review platforms, and automation frameworks. Remove the abstraction layer entirely and you don’t simplify the system—you
destabilize it.

Too many tools across the production ecosystem still depend on file semantics to function reliably.

In that framing, the file system isn’t disappearing. It’s evolving into middleware.

Werhun also noted that this layer of infrastructure is still early in its lifecycle. Technologies that present file semantics over object storage are only beginning to gain traction, and as adoption grows, more vendors are likely to emerge building solutions in that space.

Creatives Don’t Design Around Infrastructure

Architecture aside, the human factor quickly entered the conversation.

“The file system is sticking around as long as users are accustomed to using it,” said Dave Simon of Backblaze.

Simon pointed to years of experience working with media organizations and user groups. Sports teams, broadcast crews, and post-production houses expect folder hierarchies. They expect mounted volumes. They expect naming conventions that feel tangible and familiar.

Even platforms like Google Drive replicate file system views because familiarity drives productivity.

Ryan Servant of Suite Studios reinforced the point more bluntly. Creatives don’t necessarily care what’s underneath the hood. They simply want to see their files immediately, regardless of tier, location, or
lifecycle state.

And that is where the tension sharpens.

Object-native infrastructure may be architecturally elegant. But architecture rarely wins arguments on its own.

If lifecycle policies introduce latency at the wrong moment, if a file has been tiered down when an editor suddenly needs it, the user doesn’t see optimization. They see friction.

The industry is no longer optimizing solely for storage efficiency.

It is optimizing for creator delight.

The Workflow Reality Check

Theoretical architectures always look clean on diagrams. Real production environments rarely do.

Simon offered a practical scenario: imagine a field shoot generating terabytes of camera card data. In a purely cloud-object workflow, that material must first be uploaded before editing can begin.

“If Premiere starts reading S3 natively, that’s great,” Simon said. “It’s great for my business. But you still have to get the media there first.”

It wasn’t anti-object rhetoric. It was operational math.

High-performance disk tiers still solve ingest bottlenecks. Local caching still protects edit timelines. Certain transcode engines still require mount semantics to read growing files during processing.

Many production environments also continue to rely on traditional shared file systems delivered through NAS and SAN infrastructure. These architectures provide deterministic performance for demanding
workloads like high-resolution editing, finishing, and broadcast playout, and they remain a foundational part of many on-premises media workflows.

Even in five years, there will be components inside applications that rely on traditional file behavior.

Object-native does not automatically mean performance-native.

Another practical consideration came from Dave Helmly of Adobe, who noted that interacting directly with object storage introduces operational layers that traditional file systems abstract away. Accessing S3 requires credential management, client configuration, and secure key handling; processes most creative applications were never originally designed to manage internally.

Helmly also pointed to the importance of caching technologies and emerging workflow techniques like Time Addressable Media (TAMS), which allow editors to work with proxy-style representations of media
while the underlying files remain distributed across storage tiers.

These approaches help maintain timeline responsiveness while bridging the gap between object storage architectures and the real-time expectations of editing systems.

In other words, the industry isn’t simply replacing file systems with object storage.

It is building new layers that preserve the editing experience while allowing storage architectures to evolve underneath.

At the same time, object storage providers are actively working to close the performance gap. Simon noted that Backblaze has developed the ability to perform live reads of growing files directly from object
storage, allowing systems to access media even as it is still being written.

And in media production, performance still wins arguments.

Archive Isn’t Disappearing — It’s Moving Closer to Production

When the discussion turned toward archive, the existential question deepened. If object storage makes data increasingly accessible, does archive even remain a meaningful category?

Nathan Halverson of Spectra Logic brought nuance to the answer.

Object storage introduced lifecycle tiering—hot, cool, deep archive—across hybrid environments. But that flexibility increases complexity.

“Everyone says S3 is S3,” Halverson noted. “It’s a lot more complex than that.”

Retrieval policies vary. API implementations differ. Latency characteristics shift depending on tier.

What appears to be a single namespace can behave very differently depending on how lifecycle policies and storage classes are configured.

Archive isn’t vanishing.

It is becoming programmable.

It is no longer a passive vault; it is an actively orchestrated layer in the stack. And as workflows grow more distributed, that orchestration becomes more strategic, not less.

The Governance Headache

Perhaps the most forward-looking moment of the session came from the audience.

Jason Whetstone, Senior Product Development Engineer at CHESA, observed that younger media professionals increasingly think of their data as living inside applications rather than on shared file systems.

“What’s a file?” he asked, half rhetorically.

It was both humorous and revealing.

SaaS editing platforms, generative AI tools, and cloud-native collaboration systems increasingly encapsulate media within application boundaries. From a creative perspective, that feels efficient.

From a governance perspective, it creates fragmentation.

Instead of one authoritative namespace, organizations now face dozens of application-bound silos.

Servant acknowledged the tension candidly. Creatives open their preferred tool and expect their assets to be there instantly.

If governance policies move something to a lower tier, they don’t see lifecycle optimization; they see disruption.

Another audience participant pointed out an additional complication: files in professional environments rarely belong to a single application. Assets are routinely accessed by multiple tools across the production
pipeline.

“More than one app needs to access that file,” Whetstone noted.

Which is precisely why shared storage semantics remain important.

Audience member Nina Smith raised another important reminder: media workflows are rarely uniform across an organization. Editing teams, archive teams, and operations groups often have very different
requirements. Understanding who the system is truly serving is essential before designing a single unified architecture.

Where the Center of Gravity Really Lives

Late in the session, Nathan Halverson of Spectra Logic offered an insight that reframed the debate entirely. The center of gravity, he suggested, may not reside in the file system; or in object storage at all. Instead, it lives in the application layer.

Users don’t interact with storage tiers, APIs, or lifecycle engines. They interact with tools. From the user’s perspective, the application defines the experience, while the infrastructure behind it remains largely
invisible.

That perspective reorganizes the architecture of the stack. Object storage becomes the durable namespace, file systems act as performance and compatibility layers, and lifecycle engines orchestrate
movement across tiers. Applications, ultimately, define how all of those systems are experienced.

In that sense, the crown isn’t simply passing from file systems to object storage.

It’s moving upward.

So, Is the File System Dying?

No. But it may be losing its throne.

The file system is unlikely to disappear anytime soon. Too many workflows rely on its semantics, too many tools depend on its behavior, and too many users expect the familiarity of mounted volumes and directory structures.

What is changing is its role.

Rather than serving as the unquestioned foundation of media infrastructure, the file system is increasingly becoming a high-performance edge tier sitting atop object-native storage architectures.

Object storage continues to rise. Governance complexity is increasing. Applications are becoming more storage-aware, and lifecycle strategy is becoming a central architectural concern.

In that evolving stack, the file system remains essential. Not as the monarch of the infrastructure layer, but as its mediator; bridging the expectations of creative tools with the realities of modern storage systems.

And in contemporary media workflows, mediation may ultimately prove more valuable than domination.