Stats on compact block reconstructions

The examination of extra pools within nodes reveals several key insights into their nature and composition, diverging from the assumption that these pools are homogenous across the network. One primary observation is the limited size of these extra pools, which typically contain only 100 transactions by default. This small capacity suggests a fundamental difference in how extra pool transactions are treated compared to those in the mempool. Specifically, whereas mempool transactions are designed to be relayed across nodes with the aim of achieving convergence within the network, transactions within an extra pool are halted at their initial hop and are not further relayed. This mechanism inherently limits the potential for similarity between the extra pools of different nodes.

Furthermore, the diversity of extra pool contents is influenced by the unique peer-to-peer connections of each node. A node may receive numerous transactions from a peer that are subsequently rejected and relegated to its extra pool; however, another node without a direct connection to this peer would not encounter these same transactions. Consequently, the composition of an extra pool is significantly shaped by the node's network interactions and the transactional data it receives from specific peers. This aspect underscores the individualized nature of extra pools and supports the notion that they are likely to exhibit considerable variance across the network, further challenging the premise of uniformity among them.