Closed as duplicate of#10829
Description
Version
0.35
Description
@hsanjuan noticed:
the Problem is I had [Reprovider.Strategy] "all". It did not finished providing. Stored the queue on disk. Changing the strategy keeps providing what was left.
This feels like a bug. People who cant reprovide with Reprovider.Strategy=all
will not be able to benefit from alternative strategy if the node already populated the queue with everything, and there is no way to reset it.
Fix ideas
- Add
ipfs routing reprovide reset
andipfs routing reprovide reset
for manual purging of queues - Add code that detects
Reprovider.Strategy
change across restarts, and purge the reprovider queue once detected