In disaggregated mode, we currently discards the root page of an older checkpoint when a new checkpoint is written. Each checkpoint root page is assigned a unique page ID, meaning we must explicitly discard old root pages once their checkpoints become redundant (see WT-14700).

This ticket investigates the feasibility and benefits of reusing the same page ID for checkpoint root pages. If the same page ID can be reused, the new checkpoint would simply overwrite the old root page, and there would be no need for separate discard operations.

Goals:

• Investigate whether reusing the same page ID for checkpoint root pages is feasible.

• Identify any risks associated with this approach.

• Compare the potential benefits of reusing page IDs versus the current model (unique IDs with explicit discards).