|
|
Quarterly Technical Report, April 2003
Progress:
This quarter we continued our work on the Spines overlay network infrastructure
and on the Wackamole NxWay failover for servers and routers. We have also begun
exploring issues related with Domain Name Service (DNS) availability.
- The Spines overlay network infrastructure.
We designed a framework for application level, transparent reliable multicast
using the hop-by-hop reliability in Spines. The framework includes end-to-end
reliablility, congestion and flow control, and relaxed semantics over reliable
multicast that handle partitions, merges, crashes and recoveries. We started
the implementation of this framework in our overlay infrastructure.
We investigated some of the survivability aspects of Spines, both in wireless
and wired environments. We developed a mechanism of trust based on monitoring
the abnormal behaviour of overlay nodes, and an acusation system that would
eventually reroute packets to avoid untrusted nodes.
We released the first version of Spines (www.spines.org)
under a standard BSD licence.
- Highly available Domain Name Service infrastructure
The current DNS infrastructure suffers from several major drawbacks
that impact the reliability and the quality of the
provided service. Each DNS zone is served by a set of servers
organized in a single-master - multiple-slaves architecture. Under
this model, zone updates can be performed only at the master server
and they are passively propagated to the slaves through a pull
mechansim based on polling. If the master server of a zone becomes
unavailable zone updates can no longer be applied.
Furthermore, the entire infrastructure is highly dependent on
the availability of the 13 root servers. A recent denial of service
attack disabled 9 out of the 13 root servers exposing the vulnerability
of the whole system.
We have begun exploring the possibility of employing a peer zone management
system to replace the current master-slave architecture. Such a system
will maintain replicated copies of the DNS records at all the servers and
will allow for dynamic zone updates to be submitted to any peer.
Each update is propagated as soon as possible to all other servers,
reducing to a minimum the time necessary for an update to reach all
the slaves and enhancing the overall availability of the system.
Papers:
|
Reliable Communication in Overlay Networks
|
ps,
ps.gz,
pdf.
To appear in the Proceedings of the IEEE International Conference on
Dependable Systems and Networks (DSN03), San Francisco, June 2003.
Yair Amir and
Claudiu Danilov.
Reliable point-to-point communication is usually achieved in overlay
networks by applying TCP/IP on the end nodes of a connection.
This paper presents an hop-by-hop reliability approach that
considerably reduces the latency and jitter of reliable connections.
Our approach is feasible and beneficial in overlay networks that
do not have the scalability and interoperability requirements of
the global Internet.
The effects of the hop-by-hop reliability approach are quantified
in simulation as well as in practice using a newly developed
overlay network software that is fair with the external traffic
on the Internet. The experimental results show that
the overhead associated with overlay network processing at the
application level does not play an important factor compared with
the considerable gain of the approach.
|
|
N-Way Fail-Over Infrastructure for Survivable Servers and Routers. | |
To appear in the Proceedings of the IEEE International Conference on
Dependable Systems and Networks (DSN03), San Francisco, June 2003.
Yair Amir, Ryan Caudy, Ashima Munjal, Theo Schlossnagle and Ciprian Tutu.
Maintaining the availability of critical servers and routers is an important
concern for many organizations. At the lowest level, IP addresses represent the
global namespace by which services are accessible on the Internet.
We introduce Wackamole, a completely distributed software solution
based on a provably correct algorithm that negotiates the
assignment of IP addresses among the currently available servers upon
detection of faults. This reallocation ensures that at any given time
any public IP address of the server cluster is covered exactly once,
as long as at least one physical server survives the network fault.
The same technique is extended to support highly available routers.
The paper presents the design considerations,
algorithm specification and correctness proof, discusses
the practical usage for server clusters and for routers,
and evaluates the performance of the system.
|
Software:
We released the first version of Spines (www.spines.org)
under a standard BSD licence. The current version offers both best-effort and reliable
communication, obtaining better performance for reliable sessions in an overlay network setup,
compared with the end-to-end reliable communication.
Plans for Next Quarter:
Our focus for the next quarter will be on providing reliable multicast
functionality in overlay networks, and add survivabilty features to our overlay
network platform. We will continue exploring aspects related
to DNS availability.
Questions or comments to: webmaster@cnds.jhu.edu
TEL: (410) 516-5562
FAX: (410) 516-6134
|
Center for Networking and Distributed Systems
Computer Science Department
Johns Hopkins University
3400 N. Charles Street
Baltimore, MD 21218-2686
|
|
