SI-22 asks you to identify alternative sources of information for defined essential functions, and to use an alternative when the primary source is corrupted or unavailable. NIST 800-53 binds US federal systems and, by contract, some others. For everyone else it is a way to think about single points of failure in data feeds.
What it means in practice
Services act on the information they receive. If that information is corrupted, fabricated, modified or deleted, the service can fail or do harm. NIST notes that a backup source may be less precise or less accurate than the primary, yet still allow the function to continue.
Think of time sync from more than one server, a threat feed with a second provider, or a pricing service with a cached fallback.
Evidence to collect
- A list of essential functions and their primary and alternate sources.
- Configuration showing the alternate is wired in.
- Test results from cutting the primary and confirming the function continues.
- Notes on the accuracy tradeoff and who accepted it.
Common gaps
- A second source that is configured but never tested.
- Both sources sharing one upstream, so they fail together.
- No rule for when to switch back.
How to implement it
Choose two or three essential functions. Name the alternate, decide how the system detects a bad primary, and run a failover test. Record the result.
Quick check
Must the alternate source be as good as the primary?
No. The guidance accepts a less accurate source if it still lets the essential function run.
