Which design feature helps maintain cache invalidation flow during WAN outages in a multi-cluster FlexCache setup?

Which design feature helps maintain cache invalidation flow during WAN outages in a multi-cluster FlexCache setup?

• A. Inter-cluster peering redundancy
• B. Centralized origin caching
• C. Write-back caching
• D. Snapshot replication

Answer: (A)

Cache coherence across a multi-cluster FlexCache setup relies on invalidate messages traveling between the origin and all caches. When the WAN link between clusters goes down, a single inter-cluster path can break, risking stale data being served. Inter-cluster peering redundancy provides multiple, alternate communication paths for those cache invalidation messages, so the coherence signals still reach every cache even if one link fails. This ensures the invalidation flow remains intact during WAN outages and cache consistency is preserved. The other options don’t specifically guarantee continuous cache invalidation flow during WAN outages: centralizing origin caching creates a single point of access and can worsen outage impact; write-back caching affects write latency and consistency but doesn’t guarantee invalidate signals reach all caches when links fail; snapshot replication handles data copies for DR, not the real-time cache coherence signaling between clusters.

Cache coherence across a multi-cluster FlexCache setup relies on invalidate messages traveling between the origin and all caches. When the WAN link between clusters goes down, a single inter-cluster path can break, risking stale data being served. Inter-cluster peering redundancy provides multiple, alternate communication paths for those cache invalidation messages, so the coherence signals still reach every cache even if one link fails. This ensures the invalidation flow remains intact during WAN outages and cache consistency is preserved.

The other options don’t specifically guarantee continuous cache invalidation flow during WAN outages: centralizing origin caching creates a single point of access and can worsen outage impact; write-back caching affects write latency and consistency but doesn’t guarantee invalidate signals reach all caches when links fail; snapshot replication handles data copies for DR, not the real-time cache coherence signaling between clusters.