Data Archiving & Backup Guide for VFX Studios

The 3-2-1 Backup Rule

The 3-2-1 rule is the foundation of every reliable backup strategy, and it is especially critical for VFX studios. The concept is simple: maintain 3 copies of your data, stored across 2 different media types, with 1 copy kept offsite.

This rule was first articulated by photographer Peter Krogh and has since become the gold standard in data protection. For VFX studios, the stakes are particularly high. A single feature film project can generate anywhere from 50TB to 500TB of data, encompassing source plates, CG renders, compositing projects, and final deliverables. Losing any of that data mid-production can cost hundreds of thousands of dollars in re-work, and losing final deliverables after delivery can expose you to contractual liability.

Why 3-2-1 Matters for VFX

VFX data has unique characteristics that make standard backup approaches insufficient:

• Massive file sizes. EXR sequences from a single shot can be 50GB or more. A single Nuke script may reference hundreds of files across multiple directories. Traditional file-level backup tools often choke on these volumes.
• Complex project dependencies. VFX projects are not self-contained files. A Houdini scene may reference cached simulations, texture maps, and HDRIs spread across different mount points. Backing up the .hip file alone is worthless without its dependencies.
• Client contractual obligations. Most studio MSAs require you to retain final deliverables and source materials for 3-5 years after delivery. Losing this data can result in breach of contract claims.
• TPN compliance requirements. The Trusted Partner Network (TPN) assessment includes specific questions about backup procedures, disaster recovery plans, and data retention policies. Studios without documented backup strategies will fail their assessment.

Storage Media Comparison

Choosing the right storage media depends on your studio's budget, data volume, access patterns, and retention requirements. Each media type has distinct tradeoffs in capacity, speed, cost, and durability.

For most VFX studios, the optimal approach combines multiple media types. Use NVMe SSDs for active projects that require high-speed I/O (compositing, rendering), HDDs in RAID arrays or NAS for secondary storage and near-line access, and LTO tape or cloud for long-term archival. This layered approach aligns with the 3-2-1 rule while keeping costs manageable.

A Note on RAID

RAID is not a backup. RAID protects against drive failure, but it does not protect against accidental deletion, ransomware, file corruption, fire, flood, or theft. A RAID array is a single copy of your data with fault tolerance -- it is copy 1 in the 3-2-1 model, not a substitute for copies 2 and 3.

LTO Tape Workflows

LTO (Linear Tape-Open) tape remains the most cost-effective medium for long-term archival of VFX data. The latest generation, LTO-9, offers 18TB of native capacity per cartridge (up to 45TB with hardware compression), with a transfer rate of approximately 400 MB/s. At roughly $4-7 per terabyte, tape is an order of magnitude cheaper than any other storage medium for cold data.

What to Archive

Not everything needs to go to tape. A practical tape archiving strategy for VFX studios should focus on:

• Final deliverables. DPX/EXR sequences, ProRes masters, and DCP packages that were delivered to the client. These are your contractual obligations and the most critical data to preserve.
• Source plates and camera originals. The raw footage provided by production. If these are lost, they cannot be recreated.
• Completed project files. Nuke scripts, Maya scenes, Houdini files, and their associated caches and textures for completed projects. Archive these as a unit so they can be reopened if needed.
• Reference and editorial. Edit references, VFX breakdowns, client notes, and turnover packages that document the project history.

Labeling and Cataloging

A tape is only as useful as your ability to find what is on it. Every tape should be labeled with a unique identifier (e.g., AT-2026-0042) and logged in a catalog database. The catalog should record the tape ID, write date, project name, contents summary, and verification status. Many studios use dedicated media asset management (MAM) tools like CatDV or Xendata for this, but even a well-maintained spreadsheet is better than nothing.

Verification

Always verify your tape writes. After writing data to tape, read it back and compare checksums (MD5 or SHA-256) against the source. LTO drives support hardware-level verification, but software verification using tools like md5deep or hashdeep provides an additional layer of confidence. A tape backup that has not been verified is not a backup -- it is a hope.

Cloud Backup Solutions

Cloud storage provides an excellent offsite backup solution for VFX studios, but the cost model requires careful planning. Cloud providers offer multiple storage tiers, each with different pricing for storage, retrieval, and API operations. Choosing the wrong tier can result in either excessive monthly costs or prohibitively expensive retrieval fees when you actually need your data.

AWS S3 Storage Tiers

Amazon S3 is the most commonly used cloud storage platform for VFX studios. Here is how its tiers compare for archival use:

When to Use Each Tier

S3 Standard ($23/TB/mo) is appropriate for data that is actively being accessed -- for example, shared assets across geographically distributed teams, or render farm inputs that need low-latency access. Most VFX studios should minimize the amount of data in Standard tier to control costs.

S3 Infrequent Access ($12.50/TB/mo) is a good fit for projects that have wrapped but may need to be revisited within the next 6-12 months. The lower storage cost comes with a per-GB retrieval fee, but retrieval is instant, making it practical for occasional access.

S3 Glacier ($4/TB/mo) is the sweet spot for most VFX archival. Storage costs are low enough to make 100TB+ archives economically viable, and retrieval -- while not instant -- is fast enough for most non-emergency use cases. A 3-5 hour wait to pull back a project from two years ago is perfectly acceptable.

S3 Glacier Deep Archive ($1/TB/mo) is the cheapest option for data you are required to retain but are unlikely to ever need again. At $1/TB/month, storing 100TB costs only $100/month. However, retrieval is expensive ($90/TB) and slow (12-48 hours). Use this for contractual retention obligations where the data will likely never be accessed.

Backblaze B2: The Budget Alternative

Backblaze B2 offers a simpler pricing model at $6/TB/month with free egress up to 3x your stored data. For studios that find AWS pricing confusing or unpredictable, B2 is an excellent alternative. The lack of storage tiers means you pay one price regardless of access patterns, which simplifies budgeting. Backblaze also offers a partnership with Cloudflare that eliminates egress fees entirely through the Bandwidth Alliance.

Disaster Recovery Planning

A backup is only as good as your ability to restore from it. Disaster recovery (DR) planning goes beyond having copies of your data -- it defines how quickly you can resume operations after a catastrophic event and how much data you can afford to lose.

Key Concepts: RTO and RPO

Recovery Time Objective (RTO) is the maximum amount of time your studio can be down before it causes unacceptable business impact. For a VFX studio in active production with client milestones, an RTO of 24-48 hours is typical. Longer than that and you start missing deadlines and incurring penalties.

Recovery Point Objective (RPO) is the maximum amount of data loss you can tolerate, measured in time. An RPO of 4 hours means you can afford to lose up to 4 hours of work. For VFX studios, this translates to the frequency of your backup snapshots -- if you back up nightly, your RPO is 24 hours (you could lose a full day of work).

Practical Example: 50-Seat VFX Studio

Consider a mid-size VFX studio with 50 artists, 200TB of active project data on a central NAS, and 3-4 concurrent projects. Here is what a practical DR plan looks like:

• RTO: 12 hours. The studio can tolerate a half-day outage. Artists can work locally on cached files, review notes, and handle administrative tasks while systems are restored.
• RPO: 4 hours. Snapshot replication runs every 4 hours from the primary NAS to a secondary NAS in a different room (or ideally, a different building). Nightly full backups go to cloud (Glacier) and LTO tape.
• Hot spare hardware. Keep one spare server and two spare workstations ready. If the primary NAS fails, the secondary can serve as a read-only fallback while the primary is rebuilt.
• Documented runbooks. Step-by-step procedures for every failure scenario: NAS failure, switch failure, internet outage, ransomware attack, fire/flood. Each runbook should be tested quarterly.
• Annual DR drill. Once a year, simulate a full disaster recovery: shut down the primary storage, restore from backup, and measure actual RTO. Document the results and address any gaps.

TPN Requirements for DR

The TPN assessment includes several controls related to disaster recovery and business continuity. Studios pursuing TPN certification must document their backup procedures, test their DR plan regularly, and demonstrate that content can be recovered within defined timeframes. Specifically, TPN expects:

• A written disaster recovery plan that is reviewed and updated at least annually
• Regular backup testing (not just backups, but verified restores)
• Encryption of backup data, especially for offsite and cloud copies
• Access controls on backup systems equivalent to production systems
• Documented retention and destruction policies for archived content

Cost Comparison

The true cost of an archiving strategy goes beyond the sticker price of drives or cloud subscriptions. You need to factor in hardware lifecycle costs, maintenance, power and cooling, staff time, and -- critically -- the cost of retrieval when you actually need your data back.

Here is a realistic cost comparison for archiving 100TB of VFX data over 5 years:

LTO tape is the clear winner on raw cost for large archives. However, cost is not the only factor. Tape requires physical handling and storage, a cataloging system, and periodic migration to newer tape generations (roughly every 10 years). Cloud eliminates all of that operational overhead at a higher price point.

For most studios, the best approach is a combination: LTO tape for the bulk archive (completed projects that are unlikely to be accessed) and cloud for disaster recovery (a secondary offsite copy that can be retrieved without physically shipping tapes). This hybrid approach gives you the cost efficiency of tape with the accessibility of cloud.

Recommendations

There is no one-size-fits-all archiving strategy. The right approach depends on your studio's size, budget, compliance requirements, and how frequently you need to retrieve archived data. Here are our tiered recommendations:

Small Studio (5-15 Artists)

For smaller studios with limited IT resources and budgets, simplicity is key. A complex tape workflow will not be maintained consistently, and the cost of an LTO drive is harder to justify at smaller data volumes.

• Primary storage: Synology or QNAP NAS with 4-8 drives in RAID 6 (40-100TB usable)
• Offsite backup: Backblaze B2 or Wasabi for cloud backup of active projects and deliverables
• Local backup: USB 3.2 external drives (rotated weekly) for a second on-site copy
• Budget: $3,000-5,000 upfront + $200-500/month cloud

Mid-Size Studio (15-50 Artists)

At this scale, data volumes typically reach 100-500TB, and the cost savings of LTO tape become significant. Studios at this size usually have at least one dedicated IT staff member who can manage a tape workflow.

• Primary storage: Enterprise NAS or SAN (Synology, QNAP, or iXsystems TrueNAS) with 100-500TB usable
• Secondary storage: Replication to a second NAS (ideally in a separate physical location or room)
• Long-term archive: LTO-9 tape library for completed projects and deliverables
• Offsite/DR: Cloud backup (S3 Glacier or Backblaze B2) for disaster recovery copy
• Budget: $15,000-30,000 upfront + $500-2,000/month cloud

Large Studio (50+ Artists)

Large studios generate petabytes of data and require enterprise-grade archive infrastructure. At this scale, automated tape libraries, dedicated archive servers, and multi-region cloud replication become necessities rather than luxuries.

• Primary storage: Enterprise SAN with high-speed fibre channel (Isilon, Pure Storage, or NetApp) -- 500TB to multi-PB
• Near-line storage: Secondary tier for recently completed projects (HDD-based NAS or object storage)
• Tape archive: Automated LTO tape library (Quantum, Spectra Logic, or IBM) with robotic loading and MAM integration
• Cloud DR: Multi-region replication to S3 Glacier or Azure Archive with automated lifecycle policies
• Monitoring: Dedicated backup monitoring and alerting (Veeam, Commvault, or similar)
• Budget: $100,000+ upfront + $2,000-10,000/month cloud