You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
{{ message }}
Repository navigation
rsk_getStorageNodes has no cap on slots per call; its sibling caps at 64 #382
rsk_get_storage_nodes (line 245) — no cap on slots per call
rsk_get_state_nodes (line 348) — caps at MAX_STATE_KEYS = 64 (line 329),
enforced at line 359 with "At most 64 keys at a time"
Two sibling methods, one guard. The uncapped one is new and public, and it runs
on the async worker, so a single large request occupies a worker for as long as
it takes.
This is the same shape as rsk_lastAccountChange: a method whose cost is
invisible at the call site is a poor thing to expose by default. That one was
given a config gate and a bounded range for exactly this reason.
What would settle it
Apply the same MAX_STATE_KEYS bound, or state why storage slots are cheaper per
unit than state keys and pick a bound from that.
Activity