Rackspace’s involvement with OpenStack and CERN at the Large Hadron Collider surfaced again late last month when the cloud hosting provider staged a London-based gathering to discuss what, when and where its cloud hosting intelligence is being deployed.
Computer Weekly has already detailed the following case study explaining the work that has undertaken here: CERN adopts OpenStack private cloud to solve big data challenges.
Group leader of the OIS group within the IT department at CERN is Tim Bell — he’s basically the guy that looks after the tech infrastructure for CERN
Talking about the software application development work that is carried out at CERN, Bell explains that the systems today “suffer” from being embarrassingly parallel.
In parallel computing, an embarrassingly parallel workload, or embarrassingly parallel problem, is one for which little or no effort is required to separate the problem into a number of parallel tasks.
“This is essentially where High Throughput Computing (HTPC) comes into play as opposed to High Performance Computing (HPC),” explains Bell.
“That is to say, a lot of compute tasks have to be carried out, but they are all executed independently,” he added.
So is CERN worth learning from?
From a developer perspective, CERN uses OpenStack and codes some of the open IP into into its own IT stack … the end result is that some of the resultant IP is contributed back to the community but some of it isn’t.
“Everything that we identify that is of interest to the community we contribute back -- that which we deem to not be of interest we make available, but do not actually contribute back," explains Bell.
The code that runs the Large Hadron Collider is therefore available and this this gives rise to our acronym of the day...
BYO-LHC: Bring Your Own Large Hadron Collider
The (Organisation européenne pour la recherche nucléaire) isknown as CERN.