In cabal, the group somehow got synced a feed that has some missing nodes. I suspect that a number of key lookups are going through that missing node, causing hyperdb to stall out waiting for that node to arrive. This seems to have the effect that even certain old data is also now inaccessible: I think the prefix trie has to traverse the missing node in order to get to that older data. As a result, only a subset of the hyperdb is available through get.
In cabal, the group somehow got synced a feed that has some missing nodes. I suspect that a number of key lookups are going through that missing node, causing hyperdb to stall out waiting for that node to arrive. This seems to have the effect that even certain old data is also now inaccessible: I think the prefix trie has to traverse the missing node in order to get to that older data. As a result, only a subset of the hyperdb is available through
get.