MongoDB is observing server crashes when we reach the `__curhs_btree_id_to_hs_id` passing 0 `btree_id` as an argument. This has happened twice in the last month. The assertion looks meaningful, so it's worth investigating what could be leading to this state.

This function was introduced as a part of the disagg great merge, so it's very likely to be disagg related.

Reduced back trace: 

mongo::printStackTrace()
mongo::(anonymous namespace)::printErrorBlock()
abruptQuit
__GI_raise
__GI_abort
__wt_abort
__curhs_btree_id_to_hs_id
__wt_curhs_open
__wt_hs_find_upd
__wt_txn_read
__wt_btcur_next
__curfile_next
__background_compact_server
start_thread
__GI___clone