📝 Pastry protocol in peer to peer networks (8 MCQs)

📖 Explanation: The leaf set stores the set of nodes whose identifiers are numerically closest to the local node, typically the $l$ smallest and largest IDs. This allows Pastry to quickly forward messages to nearby IDs and to handle failures without consulting the full routing table.

📖 Explanation: Pastry routes messages by matching the longest prefix between the current node's ID and the destination ID. When the first digit differs, the routing table entry that corresponds to that digit is consulted, because the leaf set only handles IDs that share the prefix. Therefore the routing table directs the message toward the appropriate next hop.

📖 Explanation: Pastry builds a routing table where each row corresponds to a shared prefix length and each column to a possible digit, enabling deterministic prefix matching. In contrast, Chord’s finger table contains entries that point to nodes at distances of $2^k$ around the identifier circle, providing logarithmic hop guarantees but without explicit prefix awareness. Hence option C captures the core structural difference.

📖 Explanation: The two identifiers share the first hexadecimal digit ‘3’, so the longest common prefix length is one digit. Pastry’s routing table can forward the message to a node whose prefix matches the destination in the next digit, changing ‘A’ to ‘B’. After that step, the prefix length becomes two digits, and a second hop can resolve the final digit. Therefore the protocol guarantees at most two hops.

📖 Explanation: A larger leaf set provides more immediate neighbours whose IDs are close to the local node, thereby offering alternative next‑hops when the preferred routing table entry fails. This redundancy lowers the probability that a message cannot be forwarded because all candidates are unavailable, improving reliability. However, the size of the routing table itself does not change, so latency and bandwidth are largely unaffected.

📖 Explanation: Expanding the number of columns per routing‑table row means each node stores entries for more possible digit values, effectively increasing the branching factor. This reduces the expected number of hops because the message can often jump closer to the destination in a single step. The downside is that each node must maintain more pointers, raising memory requirements and maintenance traffic. Hence option B captures the primary trade‑off.

📖 Explanation: Pastry nodes periodically run a stabilization routine that contacts neighbors to verify and update their routing tables and leaf sets. When partitions heal, nodes exchange their current views, reconciling differences and incorporating any newly reachable identifiers. This incremental approach allows the network to converge without global flooding or restarting the bootstrap process, preserving scalability while restoring consistency. Therefore option A correctly describes the convergence mechanism.

📖 Explanation: During the join process, Y sends a join message toward its destination prefix. Each node that receives this message compares Y’s identifier with the current entry for the matching prefix. If Y is numerically closer, the node replaces the existing entry with Y, ensuring the routing table always points to the nearest node for that prefix. This update happens instantly as the join message propagates.