I'm building a workflow app that generates documents and needs low-cost archive storage later. What would you recommend?

For a workflow app that creates documents now and needs cheap long-term storage later, I’d recommend a tiered storage design:

Best-practice approach

1) Store “active” documents in primary storage

Use a database + object storage for documents that are still being edited, viewed often, or need fast access.

• Database: metadata, status, ownership, version history
• Object storage: the actual generated files
  Examples: AWS S3, Google Cloud Storage, Azure Blob Storage

2) Move older documents to archive storage automatically

Once a document is no longer actively used, transition it to a colder, cheaper storage tier.

Good low-cost archive options:

• AWS S3 Glacier Deep Archive — usually the cheapest for long retention
• AWS S3 Glacier Flexible Retrieval — cheaper than normal storage, faster retrieval than Deep Archive
• Google Cloud Archive Storage
• Azure Blob Archive Tier

If your use case is “rarely accessed, but must be retained,” Deep Archive / Archive Tier is usually the best cost saver.

---

What I’d recommend depending on access pattern

If documents are rarely needed but must be retained

Choose Deep Archive / Archive tier.

Pros:

• Lowest storage cost
• Good for compliance, legal retention, long-term records

Cons:

• Retrieval can take hours
• Retrieval fees can be non-trivial

If documents may need to be retrieved sometimes

Choose a cold storage tier with faster access.

Pros:

• Lower cost than hot storage
• Faster restore times

Cons:

• More expensive than deep archive

---

Suggested architecture

A practical setup:

1. Generate document
2. Save current file in object storage hot tier
3. Save metadata in DB
4. After a retention period, run a lifecycle rule to transition files to archive storage
5. Keep a pointer in the DB to the archived object
6. On access, restore or fetch if supported

This gives you:

• low operational complexity
• low cost for old files
• easy automation

---

Important things to plan for

Retrieval latency

Archive storage is cheap because it’s not immediately accessible. If users expect instant downloads, don’t put frequently accessed files there.

Retrieval cost

Archive tiers often charge for restores and access. If users retrieve files often, “cheap storage” may become expensive overall.

Lifecycle automation

Use automatic policies so you don’t manually move files.

File immutability / retention

If documents are records, consider:

• versioning
• object lock / WORM retention
• audit logs

---

My short recommendation

If you want the simplest low-cost path:

• Use object storage for generated files
• Set lifecycle rules to move old files to archive tier
• Pick AWS S3 Glacier Deep Archive / Azure Archive / GCS Archive depending on your cloud

If you want the best balance of cost and usability, start with:

• Standard storage for 30–90 days
• then move to cold/archive storage

If you want, I can also suggest a specific cloud/provider setup based on your stack (AWS/GCP/Azure) or help design the database + storage schema for this workflow.