Distributed Systems -Tapestry and Memory Consistency models

In Tapestry routing Nodes and objects are assigned IDs from common space via a distributed hashing. ID space typically has m =

• 160

• 128

• 16

• None

In Tapestry routing Each entry in the routing table at node is of the form---------

• (i , IP address)

• (vid ,IP address)

• (wid , IP address)

• (vid wid, IP address)

In Tapestry Surrogate routing can be used to route--------------

(a)  

In Tapestry Unique spanning tree used to route to vid is used to------------

• stores object O

• publish and locate an object

• redirect directlyto the object

In Tapestry routing At each hop and including the root node OGR , the publish message creates a pointer to the object. This is the directory info and is maintained in

• Soft-state

• Global -state

• Saved-state

In Tapestry routing On Node Insertions for any node Y on the path between a publisher of object O and the root GOR , node Y should have a (a)  

n Tapestry routing On Node Insertions If X becomes the new root of existing objects, references to those objects should

• Be constructed

• Be notified

• Lead to x

• Include in their routing tables

n Tapestry routing On Node Deletions what prevent holes in their routing tables?

• Backpointers

• Replacement entries for each level

• Routing

• None

In Tapestry routing On Node failures repairing the object location pointers, routing table is done through?

• Routing

• Redundancy

• Recovery

• Replication

In Tapestry routing the latency for each hop is expected to be ----------------, compared to that for CAN and Chord protocols.

• Large

• Maintained as same

• Small

• None