Stale bootstrap nodes will block the bucket for revalidation #280
Labels
discoveryv5
Discovery v5 related issues
Comments
kdeme
added a commit
that referenced
this issue
Nov 27, 2020
* Remove also bootstrap nodes if they fail to respond * Repopulate the routing table (with bootstrap nodes) when the routing table drops low
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Bootstrap nodes are currently never removed. Also not when pinging them fails.
The idea is here that they should be highly available (trusted) nodes that one can always fallback to.
However, if a bootstrap node is unreachable (goes stale), then this has the side-effect that it will never be put in front of the bucket and practically block revalidation of all nodes in front of that bucket (until the bucket gets split at least).
Those nodes will still be "tested" for availability by doing lookups (findNode requests) though, so it is not that terrible.
Better solution would be to remove the bootstrap nodes, but always keep them aside as fallback for when the routing table drops to
< bootstrapnodes.lenpeers.The text was updated successfully, but these errors were encountered: