If the inter-cluster peering link is temporarily unavailable, what happens to cache invalidation messages?

If the inter-cluster peering link is temporarily unavailable, what happens to cache invalidation messages?

• A. They remain valid indefinitely.
• B. They stay stale until the inter-cluster link is restored.
• C. They are refreshed automatically every few seconds.
• D. They are cleared at the next origin update.

Answer: (B)

Cache coherence across clusters relies on timely delivery of invalidation messages. If the inter-cluster peering link is temporarily unavailable, those invalidation messages cannot reach the peer cluster, so each cache continues to treat its locally stored data as valid. The entries remain stale with respect to the data in the other cluster until the link is restored and the pending invalidations can be propagated and processed. The data will then refresh as needed after connectivity returns. The other options don’t fit because there isn’t an automatic, periodic refresh of invalidations, nor a guaranteed indefinite validity, nor an automatic clearing at the next origin update without the invalidations being delivered.

Cache coherence across clusters relies on timely delivery of invalidation messages. If the inter-cluster peering link is temporarily unavailable, those invalidation messages cannot reach the peer cluster, so each cache continues to treat its locally stored data as valid. The entries remain stale with respect to the data in the other cluster until the link is restored and the pending invalidations can be propagated and processed. The data will then refresh as needed after connectivity returns. The other options don’t fit because there isn’t an automatic, periodic refresh of invalidations, nor a guaranteed indefinite validity, nor an automatic clearing at the next origin update without the invalidations being delivered.