Overlay Networks & P2P
|
|
Título del Test:
![]() Overlay Networks & P2P Descripción: Advanced Topics in Networking |



| Comentarios |
|---|
NO HAY REGISTROS |
|
Indicate the correct answers regarding overlay networks (2 correct answers). Overlay networks are implemented at the application layer. Overlay networks replace the physical Internet routing. Overlay nodes are typically end systems at the network edge. Overlay networks cannot use TCP connections. Overlay networks require changes in the underlying Internet infrastructure. Indicate the correct answers regarding P2P systems (2 correct answers). All peers have equal capabilities. A central server is required to coordinate peers. Peers can act as both clients and servers. P2P systems cannot share storage resources. P2P systems avoid self-organization. Which properties are typically associated with Peer-to-Peer systems? (2 correct answers). Higher fault tolerance than client/server systems. A single point of failure. Limited scalability. Decentralized resource control. Guaranteed authoritative information. Indicate the correct answers regarding centralized P2P systems (2 correct answers). They rely on a central index server. Content is always transferred via the central server. Peers still exchange data directly. They have no single point of failure. They cannot scale beyond a few nodes. What are characteristics of Napster’s P2P model? (2 correct answers). A central server stores an index of all shared files. All queries are flooded across the overlay. Peers connect directly to download files. Napster is a fully decentralized system. Napster uses a DHT for routing. Indicate the correct statements about Napster signaling (2 correct answers). Peers send file lists to the central server. Searches are forwarded peer-to-peer using flooding. The central server replies with a list of peers. Data is exchanged using the Napster signaling protocol. Peers never communicate directly. Which statements describe decentralized P2P networks? (2 correct answers). They contain no central entities. Any peer failure stops the network. Queries are distributed using flooding. Routing tables are fixed and global. They depend on a single bootstrap server at all times. Indicate the correct answers regarding flooding in decentralized P2P networks (2 correct answers). Flooding sends queries to all neighbors. TTL limits how far a message can propagate. Flooding guarantees optimal routing paths. TTL is increased at every hop. Flooding produces very low signaling overhead. What are properties of Gnutella 0.4? (2 correct answers). Queries are flooded across the network. A central index server stores all file metadata. Responses are routed back using GUIDs. All nodes have complete routing tables. The overlay is structured using a DHT. Which fields are included in a Gnutella message header? (2 correct answers). GnodeID. TTL. Public key. MAC address. Hops. Indicate the correct answers regarding Gnutella routing (2 correct answers). Duplicate messages are discarded using GnodeID. Queries are routed using shortest path algorithms. TTL is decreased at every hop. QUERYHIT messages are flooded like queries. PONG messages are routed back to the PING origin. Which properties describe hybrid P2P systems? (2 correct answers). They introduce super peers. They eliminate all hierarchy. Leaf nodes register content at super peers. All nodes have the same workload. They require a global central server. Indicate the correct statements about super peers (2 correct answers). They handle routing for leaf nodes. They are dynamically elected. They are permanently fixed servers. They usually have higher capacity. They cannot fail. What are characteristics of Gnutella 0.6? (2 correct answers). Leaf nodes send queries to super peers. Queries are only sent to a central server. Super peers forward queries to other super peers. All nodes use flooding. Content is exchanged directly between peers. Indicate the correct statements regarding hybrid P2P compared to pure P2P (2 correct answers). Hybrid P2P reduces signaling overhead. Hybrid P2P introduces a single point of failure. Hybrid P2P scales better than pure flooding. Hybrid P2P removes decentralization. Hybrid P2P prevents peer churn. Indicate the correct answers regarding overlay topologies (3 correct answers). Overlay links are always mapped to one physical hop. Overlay topologies can be random, structured, or hierarchical. Overlay nodes must be located in the network core. The overlay topology is independent from the physical topology. Overlay networks can use different logical structures over the same physical network. Which resources can be shared in P2P overlays? (2 correct answers). CPU cycles. IP routing tables. Storage space. Operating system kernels. Physical routers. Indicate the correct answers regarding centralized P2P drawbacks (3 correct answers). Single point of failure. High resilience to attacks. Central bottleneck. Unlimited scalability. Legal vulnerability. Which statements describe search in pure P2P systems? (2 correct answers). Searches rely on flooding. Queries are routed using a DHT. No global index is maintained. All peers know all other peers. Requests are answered only by servers. Indicate the correct answers regarding flooding overhead (1 correct answer). The number of forwarded messages grows with the number of nodes. Flooding scales with O(n). Flooding has constant overhead. Flooding is independent of network size. Flooding produces very low traffic. Which statements are correct regarding TTL? (3 correct answers). TTL limits how many hops a message can traverse. A TTL larger than the network diameter may cause redundant forwarding. TTL increases on each hop. TTL guarantees optimal routing. TTL prevents infinite message loops. Which of the following are Gnutella message types? (2 correct answers). QUERY. ROUTE_TABLE_UPDATE. SEARCH. LOGIN. PING. Indicate the correct differences between Napster and Gnutella 0.4 (3 correct answers). Napster uses a central index; Gnutella does not. Gnutella relies on flooding; Napster does not. Napster performs peer-to-peer data transfer. Both use a DHT. Both rely on super peers. Which statements describe backward routing? (2 correct answers). Responses follow the reverse path of the request. GUIDs are used to identify the origin. Requests are always broadcast. Backward routing requires global routing tables. Responses are flooded. Which methods are used to bootstrap decentralized P2P systems? (3 correct answers). Peer caches. Multicast or broadcast discovery. Central certificate authority. Web-based host lists. Static IP configuration. Which statements about load in hybrid P2P systems are correct? (3 correct answers). Super peers handle more traffic. All peers have equal workload. Load is asymmetric. Super peers route queries. Leaf nodes manage the super peer layer. Indicate the correct statements regarding routing in hybrid P2P (2 correct answers). Leaf nodes send requests to super peers. Super peers flood requests in the super peer layer. All routing is centralized. Routing is purely proactive. No TTL is used. Which statements are correct? (3 correct answers). Structured P2P uses fixed overlay connections. Unstructured P2P relies on flooding. Structured P2P offers deterministic lookup. Unstructured P2P requires a DHT. Structured overlays are random. Which factors influence super peer election? (3 correct answers). Processing power. Uptime. Bandwidth capacity. File names. Search keywords. Indicate the correct answers regarding hybrid P2P drawbacks (3 correct answers). Asymmetric load on super peers. High signaling traffic. No definitive statement if content is not found. Perfect physical topology matching. Single central authority. |





