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.”
The Latest on: Data centers
via Google News
The Latest on: Data centers
- Janesville, Madison hospitality industries showing signs of pandemic comeback according to tax dataon October 11, 2021 at 6:23 pm
Covid 19 cases, hospitalizations and deaths are declining in the US, that’s the latest from the Centers for Disease Control and Prevention. And local data reflects signs of progress, too, when it ...
- Core Scientific, City of Denton to develop 300MW blockchain data centeron October 11, 2021 at 2:25 pm
Core Scientific Holding Co. , a Bellevue, Washington-based blockchain infrastructure and digital asset mining firm, on Oct. 7 announced an agreement with t ...
- Data center provider snaps up 34 acres in Rockland to expand in New York metroon October 11, 2021 at 12:44 pm
Data center provider DataBank has acquired a 34-acre site in Rockland, New York, to expand its capacity in the New York City metro region. The Dallas-headquartered company paid owner Skae Group $23 ...
- Rowan Green to Develop $300M Hyperscale Data Center Campus in Temple, Texason October 11, 2021 at 7:47 am
TEMPLE, TEXAS — Rowan Green Data, a data center developer and infrastructure provider with offices in Houston and Portland, will develop a $300 million hyperscale campus in the Central Texas city of ...
- US EEOC sues construction company for alleged racism on Google data center construction siteon October 11, 2021 at 6:20 am
Whiting-Turner accused of subjecting Black laborers to “a racially hostile work environment” during the construction of Google's data center in Clarksville, Tennessee ...
- Scala Data Centers opens more than 100 new job positions for its operations in Latin Americaon October 11, 2021 at 6:00 am
Scala Data Centers — sustainable hyperscalable data center platform founded by DigitalBridge — digital infrastructure investment holding company of DigitalBridge Group, Inc. (NYSE: ...
- Rostelecom opens new data center in Moscowon October 10, 2021 at 10:26 pm
Russia’s national telecommunications operator Rostelecom announced that it has commissioned a new data center NORD-6 in Moscow.
- The data center industry’s green gold rush must start nowon October 9, 2021 at 12:03 am
As stakeholders are increasing their pressure on data to become green, the time for us to act is now, argues Node Pole’s Christoffer Svanberg ...
- Facebook's Massive Blackout: Here's What Data Centers Have Learned So Faron October 5, 2021 at 2:38 pm
When Facebook went down for more than six hours on Monday, the outage created a cascading series of consequences around the world.
- Facebook: We Disconnected Our Data Centers From the Interneton October 5, 2021 at 2:33 pm
Facebook says that a configuration error broke its connection with a key network backbone, disconnecting Facebook from all of data centers and leaving its DNS servers unreachable, the company said.
via Bing News