Feature #8: Maximum Clock Skew
Explore how to identify maximum clock skew in a network of routers arranged in an n-ary tree. Learn to traverse the tree with DFS, track time value differences between ancestor and descendant nodes, and understand how to optimize distributed consensus algorithms by measuring clock synchronization issues.
We'll cover the following...
Description
Our network topology consists of routers interconnected in a tree structure. Messages are forwarded in this structure from ancestor nodes to descendant nodes. A system’s clock time value is stored in tree nodes representing the routers. We want to tune a distributed consensus algorithm that is parameterized by the maximum clock skew in a forwarding path. The clock skew can be defined as the difference between the time values of two routers in a network. The messages are forwarded over our tree from ancestors to descendants or vice versa. We want to find the maximum clock skew that we can encounter while forwarding the messages in the tree.
It is established that these time values cannot be the same, and clock skew is unavoidable. ...