Slashdot Mirror
IT At the LHC — Managing a Petabyte of Data Per Second
schliz writes "iTnews in Australia has published an interview with CERN's deputy head of IT, David Foster, who explains what last month's discovery of a 'particle consistent with the Higgs Boson' means for the organization's IT department, why it needs a second 'Tier Zero' data center, and how it is using grid computing and the cloud. Quoting: 'If you were to digitize all the information from a collision in a detector, it’s about a petabyte a second or a million gigabytes per second. There is a lot of filtering of the data that occurs within the 25 nanoseconds between each bunch crossing (of protons). Each experiment operates their own trigger farm – each consisting of several thousand machines – that conduct real-time electronics within the LHC. These trigger farms decide, for example, was this set of collisions interesting? Do I keep this data or not? The non-interesting event data is discarded, the interesting events go through a second filter or trigger farm of a few thousand more computers, also on-site at the experiment. [These computers] have a bit more time to do some initial reconstruction – looking at the data to decide if it’s interesting. Out of all of this comes a data stream of some few hundred megabytes to 1Gb per second that actually gets recorded in the CERN data center, the facility we call "Tier Zero."'"
News that matters? The human race is not even able to handle itself and it wants to play with atoms.
You'll never steal my Jew gold you dirty goyim!
I read that in the Lucky Charms voice. I've got this whole weird image of anti-semitic breakfast cereals going on. Silly rabbi, Trix are for kids!
Troll standards have been slipping of late.
Kwisatz Haderach
Sell the spice to CHOAM
This Mahdi took Shaddam's Throne
"Goyim" is plural. The singular is "Goy".
It should be:
You'll never steal my Jew gold you dirty goy!
Clue: Those systems use all of that (especially redundancy for said 99.999% "fabled 5-9's" uptime), and IF you *think* it can't be "scaled upwards" OS + Software-wise?? Think again!
---
See here on SQLServer and Clustering:
http://www.google.com/search?q=%22SQLServer%22+and+%22Cluster%22&btnG=Search&sclient=psy-ab&hl=en&site=&gbv=1
or
http://www.google.com/search?q=%22SQL+Server%22+and+%22Cluster%22&sclient=psy-ab&hl=en&site=&gbv=1&spell=1
---
See here on Windows Compute Cluster Edition:
http://www.google.com/search?sclient=psy-ab&hl=en&site=&source=hp&q=%22Windows+Compute+Cluster%22&btnG=Search&gbv=1&sei=q-8bUJf_HeHj0QGq4YHwAg
---
"Nuff said", as the saying goes...
Especially since of what I noted in Windows Compute Cluster Edition Server (and in the case of business, SQLServer can handle THAT much too) & especially if you distribute the data across many systems with a good solid programmatic algorithm/engine...
Plus, I've actually done that kind of work (both database, OS, & programmatic architecture as well as inputs into the hardware design required with network topology too), & for some of the companies noted in my last post, using Borland Delphi &/or Microsoft Visual C++ as the language used (for performance we went with non-interpreted languages & either Oracle OR SQLServer)!
So - have you done the same? Somehow, I doubt it...
---
"It doesn't matter what OS you use or how big your data is, if you work carefully (don't pull the wrong cable or drive) and follow The Secret, you can have 100% uptime." - by Anonymous Coward on Friday August 03, @11:18AM (#40868427)
Uhm, did I state otherwise? No. You should read up on some of the case studies on the examples I noted, they use that.
* By the way? You OMITTED power redundancy protection, in UPS (uninterruptable power supplies)...
APK
P.S.=> So, what so-called "point" were you *trying* (vainly) to make here? That you can stalk/troll/harass me by ac posts?? You proved that much... thank-you I suppose!
... apk