Resilient overlay networks
read more
Citations
A delay-tolerant network architecture for challenged internets
Software-Defined Networking: A Comprehensive Survey
A taxonomy of DDoS attack and DDoS defense mechanisms
Wide-area cooperative storage with CFS
Scalable application layer multicast
References
A Border Gateway Protocol 4 (BGP-4)
A case for end system multicast
Resilient overlay networks
End-to-end Internet packet dynamics
Overcast: reliable multicasting with on overlay network
Related Papers (5)
Frequently Asked Questions (18)
Q2. What have the authors stated for future works in "Resilient overlay networks" ?
Understanding the interactions between them and investigating routing stability in an Internet with many RONs is an area for future work.
Q3. What is the importance of avoiding bad paths?
From the standpoint of improving the reliability of path selection in the face of performance failures, avoiding bad paths is more important than optimizing to eliminate small throughput differences between paths.
Q4. How long does it take to detect and route around a path failure?
Their implementation takes 18 seconds, on average, to detect and route around a path failure and is able to do so in the face of an active denial-of-service attack on a path.
Q5. How much hysteresis is used to predict the last good route?
Based on an analysis of 5000 snapshots from a RON node’s link-state table, the authors chose to apply a simple 5% hysteresis bonus to the “last good” route for the three metrics.
Q6. Why do the authors think RONs are well-suited to providing fine-grained policy routing?
Because RONs will typically run on relatively powerful end-points, the authors believe they are well-suited to providing fine-grained policy routing.
Q7. How long does it take to converge to a new valid route?
BGP-4 takes a long time, on the order of several minutes, to converge to a new valid route after a link failure causes an outage [12].
Q8. What is the common way to achieve fault-tolerant network connectivity?
An oft-cited “solution” to achieving fault-tolerant network connectivity for a small- or medium-sized customer is to multi-home, advertising a customer network through multiple ISPs.
Q9. What is the policy tag used to demultiplex the packet?
If the packet is destined for the local node, the forwarder uses the packet type field to demultiplex the packet to the RON client.
Q10. How long does it take to detect a failed path?
The time to detect a failed path suggests that passive monitoring of in-use linkswill improve the single-virtual-link failure recovery case considerably, since the traffic flowing on the virtual link can be treated as “probes.
Q11. How many seconds is the next probe packet sent?
Their implementation uses the following values:12 seconds ) 3 seconds ) 14 secondsWhen a probe loss occurs, the next probe packet is sent immediately, up to a maximum of 3 more “quick” probes.
Q12. What is the maximum error in the one-way loss rate estimate?
Assuming that losses are independent on the two paths and , then the maximum absolute error in the one-way loss rate estimate occurs when all of the loss is in only one direction, resulting in an error equal to# .
Q13. How many intermediate nodes were involved in the shortest path?
In addition, the remaining @ of the time when RON’s overlay routing was involved, the shortest path involved only one intermediate node essentially all the time: about 98%.
Q14. What is the result of the Internet’s vulnerability to router and link faults?
The result is that today’s Internet is vulnerable to router and link faults, configuration errors, and malice—hardly a week goes by without some serious problem affecting the connectivity provided by one or more Internet Service Providers (ISPs) [15].
Q15. How long does it take to detect and recover from a fault?
Their implementation takes 18 seconds, on average, to detect and recover from a fault, significantly better than the several minutes taken by BGP-4.
Q16. What is the way to limit the size of the routing tables?
To limit the size of their routing tables, many ISPs will not accept routing announcements for fewer than 8192 contiguous addresses (a “/19” netblock).
Q17. How many times did the authors take samples from a MByte bulk transfer?
The authors also took 8,855 throughput samples from 1 MByte bulk transfers (or 30 seconds, whichever came sooner), recording the time at which each power of two’s worth of data was sent and the duration of the transfer.
Q18. What is the construction of a shortest-paths algorithm?
This construction applies to multi-hop or single-hop indirection; because a standard shortest-paths algorithm may not work for all metrics (specifically, TCP throughput), their implementation, described in Section 5.2, is specific to single-hop indirection.