Use this page to choose a backup strategy that matches your custody policy and the amount of coordination you can realistically maintain.

For your first successful drill:

• use one small test file
• do not shard yet
• recover it immediately
• compare the restored bytes with the original

Complex custody schemes are only useful after the simple path has already worked.

Practical defaults:

• start with non-sharded for the first drill
• use 2-of-3 when one person must not recover alone
• use dual sharding only when policy requires it

These are different artifact classes. Treat them that way in storage and custody.

• Keep main documents, recovery documents, and shards in separate custody paths.
• Keep one independent non-Ethernity backup path.
• Record which artifact set belongs to which source and date.
• Do not store every recovery input in the same folder forever and call that a custody policy.

If the source data changes later:

• use ethernity extend to append changes to an existing unsealed root
• use ethernity compact when you want to flatten a root plus extensions into a new standalone set

Practical rule:

• use extend for normal updates
• use compact when the chain has done its job and you want a fresh root to carry forward

For the full operator flow, use Extension Workflow.

• 2-of-3 is a practical baseline for small teams.
• 3-of-5 is more resilient, but slower to assemble under pressure.
• Test the minimum quorum, not just the full set.
• Avoid thresholds your team cannot actually meet during an incident.

• Starting with complex sharding before a simple first drill.
• Storing all artifact classes together.
• Never testing minimum-quorum recovery.
• Picking a threshold that sounds good on paper but fails in practice.
• Treating printed artifacts as automatically safe without a custody plan.

• Backup Playbooks
• Extension Workflow
• Recovery Workflow
• Security and Trust Model
• Command Cheatsheet