A blockchain dashboard stops updating, but another explorer continues showing new blocks. One possible explanation is a rate limit at the dashboard’s RPC provider: the service through which the application requests blockchain information. That is a different failure from the blockchain itself stopping.
As a concrete example, Alchemy’s throughput documentation describes HTTP 429 responses when requests exceed its throughput capacity. Its limits and response formats are specific to that service. Another provider may meter requests differently, so inspect its documentation and the actual error rather than assuming every stalled screen has the same cause.
Separate the screen from its data connection
Imagine a dashboard that refreshes balances, token details and recent activity together. Several panels can stop refreshing when their shared connection is limited. A number still visible on screen may be the last successful response, not a newly verified balance.
A useful interface therefore shows when data was last updated and whether a refresh failed. Replacing a failed balance request with zero would communicate something the application has not established. Keeping the previous value clearly marked as stale is easier for a reader to interpret.
Investigate the request pattern
For an operator, record the time, endpoint, response status and affected request type without exposing API keys. Check the provider’s usage view and service status. One error message is evidence about that request; it is not proof of a chain-wide outage.
Use the provider’s recommended retry behaviour, including any retry delay it supplies. An immediate loop of repeated requests can add load. Alchemy discusses backing off retries in its throughput guidance.
For users, repeatedly pressing a transaction button is a poor way to diagnose a display problem. A stale dashboard does not by itself establish whether an earlier submission succeeded. Check the recorded transaction identifier through a trusted independent view before deciding what to do next. Request limits, displayed data and transaction outcomes are three questions worth keeping separate.