Right now if node does not respond to same audit request 4 times in a row, it will be disqualified. However, that node might still have the data and reason it did not reply to audit could be because it was having technical difficulties or was overloaded.
In order to avoid disqualification of healthy nodes due to temporary troubles, could we allow a node to “catch-up” on failed audits, proving that it still holds data and move it back to vetted mode instead of having it stay permanently disqualified.
if you fail only one audit because of no response 4 times, you’re not disqualified. Only if that happens a lot.
If your node isn’t able to respond to audits, which are very tiny requests after 4 tries on a consistent basis. How can the satellite trust your node will actually respond on time to requested pieces. Having the data is one thing, but if it can’t be downloaded back from your node, your node is useless.
That said, the new concurrency limit and database optimizations should probably fix the issue of your node getting over loaded. So the scenario you’re describing is not likely to happen anymore to begin with.