MODULE 4 Flashcards
Name:
Denotation to an entity in a distributed system
Identifier:
A name with properties
Access point:
Entities named by means of an address
Pure name:
Random string that has no meaning and is only used for comparisons
Broadcasting:
Using an entity’s ID to request the entity to return its address
Identifier properties:
- An identifier refers to only one entity
- An entity can only be referred to by one identifier
- An identifier always refers to the same entity
Chord organization:
- Each node has a random identifier
- Each entity has a unique key
- An entity is assigned to the node with the smallest ID that is greater or equal to the entity’s key (succ(k))
ARP’s job:
Finding out the MAC address associated with an IP address
HLS stands for _____
Hierarchical Location Services
Geographical scalability problems require separate chain reduction mechanisms because:
- Long chains aren’t fault tolerant
- There’s an increased network latency at dereferencing
ARP stands for _____
Address Resolution Protocol
Chord:
Nodes organized in a ring where each node and entity have unique identifiers in order for nodes to store these entities using their identifiers
HLS organization:
- Address of an entity is stored in a leaf or intermediate node
- Intermediate nodes store a pointer to a child if the child’s subtree contains the entity’s address
- Root knows all entities
HLS:
A search tree with networks divided into domains, and each domain is represented by a directory
HLS lookup operation:
- Start at local leaf
- If the node knows about the entity, follow it’s downward pointer, else go up
- Upward lookup stops at root
HLS insert operation:
- Keep forwarding request till a node that knows about the entity is reached
- Create a chain of pointers to the leaf
Self-certifying name:
Flat names for entities that include a derived value from the entity itself
HLS scaling steps:
- Choose physical servers for the logical server per entity
- Map entities to physical servers in a way that distributes the load of records
Approaches to trust that name-resolution returns what’s associated with a flat name:
- Secure identifier-to-entity associations
- Secure name-resolution process
Naming graph:
A graph where leaves represent named entities
Directory node:
An entity that refers to other nodes
Mounting:
Associating a node identifier of another name space with a node in current name space
Node attributes:
- Entity type
- Entity identifier
- Entity address
- Nicknames
Directory node contains:
- Table of (node identifier; edge label) pairs
- Attributes
Closure mechanism:
A mechanism that selects the context to start with for name resolution
Edge label:
Incoming edges to nodes
Foreign name space:
The name space that needs to be accessed for mounting
Mount point:
The node in current name space for mounting
Implicit location dependency:
Mapping nodes to servers that can be located anywhere
Mounting point:
The node in foreign name space that continues the name resolution for mounting
DNS stands for _____
Domain Name Service
Requirements for mounting across a network:
- Access protocol name
- Server name
- Mounting point name
DNS:
Hierarchical name space with each node having exactly one edge label
Domain:
Subtree
Domain name:
A path name to the domain’s root
Securing a DNS:
Grouping records of the same type into a signed set per zone
Zone-signing key:
The public key of the secret key used for signing a set of records
Process for trusting the signatures:
- Zone-signing keys are grouped into another set that is signed by another secret key
- Hash of key-signing key is signed by the parent zone
Key-signing key:
The public key of the secret key used for signing a set of zone-signing keys
LDAP stands for _____
Lightweight Directory Access Protocol
Drawbacks of distributed index:
- A query with n attributes requires contacting n servers
- No support for range queries
Space-filling curve steps:
- Map n-dimensional space of n attributes into a one dimension
- Hash values to distribute the 1-dimensional space among index servers
