New system could reduce data-transmission delays across server farms by 99.6 percent
Big websites usually maintain their own "data centers," banks of tens or even hundreds of thousands of servers, all passing data back and forth to field users' requests. Like any big, decentralized network, data centers are prone to congestion:
Packets of data arriving at the same router at the same time are put in a queue, and if the queues get too long, packets can be delayed.
At the annual conference of the ACM Special Interest Group on Data Communication, in August, MIT researchers will present a new network-management system that, in experiments, reduced the average queue length of routers in a Facebook data center by 99.6 percent — virtually doing away with queues.
When network traffic was heavy, the average latency — the delay between the request for an item of information and its arrival — shrank nearly as much, from 3.56 microseconds to 0.23 microseconds.
Like the Internet, most data centers use decentralized communication protocols: Each node in the network decides, based on its own limited observations, how rapidly to send data and which adjacent node to send it to. Decentralized protocols have the advantage of an ability to handle communication over large networks with little administrative oversight.
The MIT system, dubbed Fastpass, instead relies on a central server called an "arbiter" to decide which nodes in the network may send data to which others during which periods of time. "It's not obvious that this is a good idea," says Hari Balakrishnan, the Fujitsu Professor in Electrical Engineering and Computer Science and one of the paper's coauthors.
With Fastpass, a node that wishes to transmit data first issues a request to the arbiter and receives a routing assignment in return. "If you have to pay these maybe 40 microseconds to go to the arbiter, can you really gain much from the whole scheme?" says Jonathan Perry, a graduate student in electrical engineering and computer science (EECS) and another of the paper's authors. "Surprisingly, you can."
Division of labor
Balakrishnan and Perry are joined on the paper by Amy Ousterhout, another graduate student in EECS; Devavrat Shah, the Jamieson Associate Professor of Electrical Engineering and Computer Science; and Hans Fugal of Facebook.
The researchers' experiments indicate that an arbiter with eight cores, or processing units, can keep up with a network transmitting 2.2 terabits of data per second. That's the equivalent of a 2,000-server data center with gigabit-per-second connections transmitting at full bore all the time.
"This paper is not intended to show that you can build this in the world's largest data centers today," Balakrishnan says. "But the question as to whether a more scalable centralized system can be built, we think the answer is yes."
Moreover, "the fact that it's two terabits per second on an eight-core machine is remarkable," Balakrishnan says. "That could have been 200 gigabits per second without the cleverness of the engineering."
The key to Fastpass's efficiency is a technique for splitting up the task of assigning transmission times so that it can be performed in parallel on separate cores. The problem, Balakrishnan says, is one of matching source and destination servers for each time slot.
"If you were asked to parallelize the problem of constructing these matchings," he says, "you would normally try to divide the source-destination pairs into different groups and put this group on one core, this group on another core, and come up with these iterative rounds. This system doesn't do any of that."
Instead, Fastpass assigns each core its own time slot, and the core with the first slot scrolls through the complete list of pending transmission requests. Each time it comes across a pair of servers, neither of which has received an assignment, it schedules them for its slot. All other requests involving either the source or the destination are simply passed on to the next core, which repeats the process with the next time slot. Each core thus receives a slightly attenuated version of the list the previous core analyzed.
Today, to avoid latencies in their networks, most data center operators simply sink more money into them. Fastpass "would reduce the administrative cost and equipment costs and pain and suffering to provide good service to the users," Balakrishnan says. "That allows you to satisfy many more users with the money you would have spent otherwise."
Networks are typically evaluated according to two measures: latency, or the time a single packet of data takes to traverse the network, and throughput, or the total amount of data that can pass through the network in a given interval.
Written by Larry Hardesty, MIT News Office
Andrew Carleen | Eurek Alert!
Five developments for improved data exploitation
19.04.2017 | Deutsches Forschungszentrum für Künstliche Intelligenz GmbH, DFKI
Smart Manual Workstations Deliver More Flexible Production
04.04.2017 | Deutsches Forschungszentrum für Künstliche Intelligenz GmbH, DFKI
More and more automobile companies are focusing on body parts made of carbon fiber reinforced plastics (CFRP). However, manufacturing and repair costs must be further reduced in order to make CFRP more economical in use. Together with the Volkswagen AG and five other partners in the project HolQueSt 3D, the Laser Zentrum Hannover e.V. (LZH) has developed laser processes for the automatic trimming, drilling and repair of three-dimensional components.
Automated manufacturing processes are the basis for ultimately establishing the series production of CFRP components. In the project HolQueSt 3D, the LZH has...
Reflecting the structure of composites found in nature and the ancient world, researchers at the University of Illinois at Urbana-Champaign have synthesized thin carbon nanotube (CNT) textiles that exhibit both high electrical conductivity and a level of toughness that is about fifty times higher than copper films, currently used in electronics.
"The structural robustness of thin metal films has significant importance for the reliable operation of smart skin and flexible electronics including...
The nearby, giant radio galaxy M87 hosts a supermassive black hole (BH) and is well-known for its bright jet dominating the spectrum over ten orders of magnitude in frequency. Due to its proximity, jet prominence, and the large black hole mass, M87 is the best laboratory for investigating the formation, acceleration, and collimation of relativistic jets. A research team led by Silke Britzen from the Max Planck Institute for Radio Astronomy in Bonn, Germany, has found strong indication for turbulent processes connecting the accretion disk and the jet of that galaxy providing insights into the longstanding problem of the origin of astrophysical jets.
Supermassive black holes form some of the most enigmatic phenomena in astrophysics. Their enormous energy output is supposed to be generated by the...
The probability to find a certain number of photons inside a laser pulse usually corresponds to a classical distribution of independent events, the so-called...
Microprocessors based on atomically thin materials hold the promise of the evolution of traditional processors as well as new applications in the field of flexible electronics. Now, a TU Wien research team led by Thomas Müller has made a breakthrough in this field as part of an ongoing research project.
Two-dimensional materials, or 2D materials for short, are extremely versatile, although – or often more precisely because – they are made up of just one or a...
20.04.2017 | Event News
18.04.2017 | Event News
03.04.2017 | Event News
24.04.2017 | Life Sciences
24.04.2017 | Earth Sciences
24.04.2017 | Machine Engineering