CERN Collider To Trigger a Data Deluge

slashthedot sends us to High Productivity Computing Wire for a look at the effort to beef up computing and communications infrastructure at a number of US universities in preparation for the data deluge anticipated later this year from two experiments coming online at CERN. The collider will smash protons together hoping to catch a glimpse of the subatomic particles that are thought to have last been seen at the Big Bang. From the article: "The world's largest science experiment, a physics experiment designed to determine the nature of matter, will produce a mountain of data. And because the world's physicists cannot move to the mountain, an army of computer research scientists is preparing to move the mountain to the physicists... The CERN collider will begin producing data in November, and from the trillions of collisions of protons it will generate 15 petabytes of data per year... [This] would be the equivalent of all of the information in all of the university libraries in the United States seven times over. It would be the equivalent of 22 Internets, or more than 1,000 Libraries of Congress. And there is no search function."

Comments

OT: The size of the internet

[chriss]

Okay, the Library of Congress has been estimated to contain about 10 Terabyte, so I buy the 1000 * LoC = 15 Petabyte. But archive.org alone expanded its storage capacity to 1 Petabyte in 2005, so the CERN is not going to generate anything near "22 Internet" (whatever that might be). This estimate from 2002 calculates the size of the internet as about 530 Exabyte, 440 Exabyte of which are email, 157 Petabyte for the "surface web"

Re:OT: The size of the internet

[Anonymous Coward]

We are from NASA, and would like to offer you a job in mission planning.

No Search Function

[tacocat]

Google it?

If Google is so awesome, maybe they can put their money where there mouth is and do something commendable. Of course, they'll probably have a hard time turning this data into marketing material.

Re:No Search Function

[scheme]

The problem is less that there is no search function (with digital data all you're doing is matching one pattern to another), the problem is more that you don't know exactly what you are searching for!
My guess is that they are looking for anomalies within the data that would indicate the presence of one of these subatomic particles. My guess furthermore is that once they get enough data analyzed they will be able to form a model to base a search function around.
That or the summary lies (wouldn't be the first time) and in fact they know exactly what they are searching for, and they have a search function, but of course someone has to look at the output of those functions to determine what impact they have on their model/ideas.

For a lot of the physics, the researchers know what they are looking for. For example, with the Higgs boson, theories constrain the decay and production to certain channels that have characteristic signatures. So they would be looking for events that have a muon at a certain energy with a hadron jet with another given energy coming off x degrees away and so on. There have been monte carlo simulations and other calculations done to predict what the interesting events should look like using various different theories. Of course there maybe interesting events that pop up that no one has predicted but everyone has a fairly good idea of what the expected events should look like.

Re:No Search Function

[Benson Arizona]

Buy Higgs Boson now at e-bay.com

Buy books about Bosons at Amazon.com

60%

[Alsee]

The CERN collider will begin producing data in November, and from the trillions of collisions of protons it will generate 15 petabytes of data per year... [This] would be the equivalent of all of the information in all of the university libraries in the United States seven times over. It would be the equivalent of 22 Internets, or more than 1,000 Libraries of Congress. And there is no search function.

And 60% of it will be porn.

-

Re:60%

[carpe_noctem]

mmmm... particle porn!

See the hottest collisions on the web! Watch as innocent particles get ripped apart, revealing their inner quarks! See protons get exploited and penetrated in their luscious gluons!

Re:60%

[lexarius]

Talk like that gives me a large hadron.

Never mind the data

[simong]

What about the backups?

Neutrinos

[MichaelSmith]

I hope they're planning on running their own fiber optic line across the Atlantic

You know with the right sort of particle accelerator you could send messages straight through the Earth and save a heap of latency.

Re:Neutrinos

[Dunbal]

You know with the right sort of particle accelerator you could send messages straight through the Earth and save a heap of latency.

It's called the "Death Star" project, and we've been having a hell of a time with the receiver...

22 Internets per year?

[UnHolier than ever]

Would that be 0.84 Internet per forthnight? Or 1 kiloLibrary per Congress session? How much in tubes?

Re:Is there a danger or isn't there?

[SamSim]

all the particle colliders of the most recent generation (like the Tevatron at Fermilab or the Relativistic Heavy Ion collider in New York) have the capability (if certain theoretical models are accurate enough) to generate very tiny (around nine millimeters), but stable black holes (though the probability is extremely low)

Well, yeah, but the probability is about the same as that of you generating a small black hole by clapping your hands together really hard.

All pages are identical

[Laxator2]

The main difference between the LHC data and the Internet is that all that 15 PB of data will come in a standard format, so a search is much easier to perform. In fact most of the search will consist on discarding non-interesting stuff while attempting to identify the very rare events that may show indications of new particles (Higgs for example). The Internet is a lot more diverse, the variety of information dwarfs the limited number of patterns LHC is looking for, so "no search available" for LHC data sounds more like "no search needed".

I predict the end of the universe

[jamesh]

This is really bad news. By defining the amount of data in LoC's, they leave themselves open to a huge exploit... If the LoC ever includes this data, then there will be a recursive loop of definitions and the LoC will expand to fill the universe.

Okay... maybe not, but if they ever did put this data in the LoC, the effort required to re-factor all the LoC based measurements would bankrupt the world. And the confusion that goes on while this re-factoring is happening will surely crash at least one probe into Mars, where the English have used the new LoC units and the Americans will have used the old LoC units.

Re:I predict the end of the universe

[TapeCutter]

It seems the metric LoC = 10TB. If that is so then an LoC is no longer based on a physical library but has rather been redefined based on a more basic unit of information, (ie: the byte). This sort of thing has happened before, the standard time unit (second) is no longer based on the earth's rotation, rather it is based on some esoteric (but very stable) feature of cesium atoms.

IMHO: This is a GoodThing(TM), it could mean the LoC is well on it's way to becoming an accepted SI unit. :)

Re:Too much for the 'Net

[bockelboy]

That's 4Gbps AVERAGE, meaning it's much below the peak rate. That's also the raw data stream, not accounting for site X in the US wanting to read reconstructed data from site Y in Europe.

LHC-related experiments will eventually have 70 Gbps of private fibers across the atlantic (Most NY -> Geneva, but at least 10Gbps NY -> Amsterdam), and at least 10 Gbps across the Pacific.

For what it's worth, here's the current transfer rates for one LHC experiment You'll notice that there's one site, Nebraska (my site), which averages 3.2 Gbps over the last day. That's a Tier 2 site - meaning it won't even recieve the raw data, just reconstructed data.

Our peak is designed to be 200TB / week (2.6Gbps averaged over a whole week). That's one site out of 30 Tier 2 sites and 7 Tier 1 sites (each Tier 1 should be about 4-times as big as a Tier 2).

Of course, the network backbone work has been progressing for years. It's to the point where Abilene, the current I2 network, rarely is at 50% capacity.

The network part is easy; it's a function of buying the right equipment and hiring smart people. The extremely hard part is putting disk servers in place that can handle the load. When we went from OC-12 (622 Mbps) to OC-192 (~10Gbps), we had RAIDs crash because we wrote at 2Gbps on some servers for days at a time. Try building up such a system without the budget to buy high-end Fiber Channel equipment too!

And yes, I am on a development team that works to provide data transfer services for the CMS experiment.

Re:Never underestimate the bandwidth of a 747

[Dunbal]

If they had an A380 (Airbus for teh win ;-)) worth of hard drives installed and ready to tap data, they would not need to move all that data.

I'm sorry, how much is that in Cessna 172's again?

Re:Remember

[bockelboy]

I do work with one of the LCG projects, so let me share some of my personal opinions with you (all this info is mostly available on the web, if you can find it. We keep no secrets.).

I don't think CERN has HTTP/FTP servers right on a OC Internet backbone, or the server structure (think magnitudes greater than Google's) to drive the data.

Oh yes we do. You are right though - buying network bandwidth is a lot more straightforward than building an disk / server infrastructure to handle all the data. It's difficult, but being accomplished.

I think total - transatlantic fiber plus the European equivalent of Internet2 - bandwidth to CERN will amount to 100 Gbps - about 10 OC-192s. Universities buy into private global fiber networks, which are independent of the public internet.

We then use gridFTP as a transport, which is basically PKI-protected FTP which transfers in N many parallel TCP streams. Then, we use a protocol called SRM to control the gridFTP transfers and (well, the CMS experiment) uses a higher-level application called PhEDEx to control worldwide data movement. Right now, PhEDEx directs about 8-10 Gbps worldwide, and we aren't "doing anything" big.

GridFTP is a fairly effective protocol. I can get near-line speed - 2Gbps from a channel bonded RAID device. Locally, we've been buying large RAIDs - 30TB a box, building up to 200TB this fall. Some sites take a more "clustered" approach - they put a few 500-750 GB drives in each of the cluster's worker nodes, and build up to 200TB that way. Costs are lower, but you have to keep 2 copies of each file in the cluster, plus have the headache of swapping out drives. Of course, I like our method better. In addition, larger, T1 sites have a few petabytes in tape silos.

Funding agencies don't just throw money into projects for years at a time, then wait for results. Two years ago, we did a test at 25% of the turn-on "complexity" (in terms of jobs run and data movement). Last year, we increased that to 50% complexity. Toward the end of this summer, we will have a challenge called CSA07 which should be between 75-100% complexity. Finally, turn-on should be around November this year.

This is a multi-billion dollar project which has been under development for 10-15 years. We've been doing lots and lots of careful planning.

Re:Never underestimate the bandwidth of a 747

[fbjon]

We obviously want to use maximum storage per HD weight, which is currently the Hitachi Deskstar 7K1000, we would have 1,000,000,000,000 bits per a maximum of 700 grams.

Using the maximum payload weight of an A380F (freighter model), we get with Google calc: (152 400 kg / 700 grams) * 1Tbytes = 193.36913 petabytes, which is 12.8912753 years worth of CERN CMS data over a maximum distance of 5,600 nautical miles.

The maximum useful load of a Cessna 172 is 371 kg, which gives a meager 0.0313823042 years worth of data over a maximum distance of 687 nm.

The raw distance between CERN and Purdue University (not including distances to airports and such) is about 3838 nm, well within range of the A380F. The Cessna 172 falls into the ground/ocean long before that however. Since there's no air-refueling option for the Cessna, the plan calls for a fleet of at least 179 Cessna 172's constantly working in relay, just to keep up with the data production rate!

So, to answer your question: If you want the same leisurely pace of using one A380F, you'll need a massive 2148 Cessnas flying for a full year, every 12 years (the total weight of which is equivalent to 531 A380F's, which should tell you something about the efficiency of said plan).