As the UK seemed to be in the grip of an extended winter the GridPP collaboration headed north to have the first collaboration meeting of 2013. While the management board had a face to face meeting on the Monday afternoon the rest of the people working on the project were en route to begin the meeting properly on Tuesday morning.

It was almost 6 years to the day since GridPP had last decamped en masse to Glasgow and the welcome was as warm as it had been that time. The meeting was opened by Prof. Steve Beaumont, the Vice Principal for Research & Enterprise at the university, followed by GridPP’s project leader David Britton giving the state of the union and setting the scene for the next two days.

The rest of Monday was dedicated to the discussions and presentations on the project’s cloud activities. With these still in their early stages, the talks covered a vast range of topics that will need to be considered. Of particular interest was Andrew McNab’s talk on LHCb’s use of clouds and virtual machines. He introduced a new idea into the mix, “vacuum computing”, as a counterpoint to the grids and clouds paradigms (see slide #5). As the day drew to a close a lot of ideas and discussion points were hammered out, with the session being led by Dave Colling from Imperial College London, Ian Collier from the Rutherford Appleton Laboratory and Jeremy Coles of the University of Cambridge.

The end of the first day could only mean one thing, the collaboration dinner. Everyone headed into the Trades Hall in the centre of Glasgow city. After a great day of work the meal was a chance to unwind and catchup, an opportunity taken by everyone there.

Tuesday morning saw a slight change of tack as the issue of “impact” dominated the early part of the day. GridPP has been up and running for over a decade now and its effects on particle physics and the LHC are undeniable. However it needs to work on maximising not only the benefits to other disciplines and society in general but making sure they shout about it. There were great contributions from Jamie Coleman of the TechCube in Edinburgh and Tom Whyntie of CERN@school. Both talks prompting lots of discussion and questions.

After coffee was an open discussion led by Neasan O’Neill, GridPP’s dissemination officer, on how GridPP can improve what it does to increase its impact work but also how it promotes these activities. And finally there was just time for the four Tier2s to give an update on their activities since the last meeting.

As everyone headed off for trains and planes they could reflect on a busy meeting with some very interesting actions and ideas to take home.

All of the talks from the meeting can be found on the meeting page for GridPP30.

The grid Hierarchy

Grid computing has created its own community of users and providers since its beginnings almost 10 years ago. As the grid matures and more researchers come on board to use it, the amount of resources available on the infrastructure also needs to grow. One solution to this is through shared facilities, like at Edinburgh. The new Sussex site is also one of these shared facilities, created in collaboration between their experimental particle physics group (EPP) and the university’s IT services (ITS). As these systems are not dedicated to grid work they don’t fit into the normal grid hierarchy (see image, left) and are referred to as Tier 3s.

The University of Sussex has been involved in GridPP since the very beginning but has not been able to provide much computing resources to the collaboration. However when EPP started talking to ITS about how they could accommodate the needs of the physics group, the solution was to provide a system for the entire campus, which would use the expertise in both groups for a common good.

“For doing analysis in any modern experiments you need to have a lot of computing resources at your disposal, very fast CPUs with high memory and a lot of disk space. Also you need a lot of bandwidth to be able to pull the data from the grid and/or submit your jobs to the grid.” says Dr Fabrizio Salvatore, from EPP “Once the resources for both ITS and EPP were in place, it was obvious for all of us that the best solution was going to be to pursue this together. From day one we have been in contact and discussed how we could accommodate the needs of EPP and ITS. This of course has meant that we benefited from each other’s expertise and the various links in the grid community to end up with the excellent resource we now have.”

The ATLAS team at the University of Sussex, Fabrizio Salvatore second from the right

Once the two groups had money and a plan of action they started the tendering process. It wasn’t the most straight forward of processes, when most GridPP sites buy hardware they are just adding it to an existing system, starting from scratch is a different story. Salvatore had previously been at Royal Holloway, University of London who host some of GridPP’s Tier 2 resources. “For me, the challenge has been the process of choosing the right hardware, the integrators for the cluster building, the parallel file system, etc.” explains Fabrizio “This would not have been possible without the help from ITS, especially Jeremy Maris, who worked on this almost full time for 3 months”.

Once the hardware was installed in the new state-of-the-art data centre on the Sussex campus it is was only the start for the EPP sys admin Emyr James and ITS’s Jeremy Maris and Tom Armour; now they had to create a grid cluster. After a lot of work they now have EMI2 running on the machines and installed the ATLAS software. The ATLAS researchers had been testing the setup without using the grid tools for a few months. “It has been remarkably easy to get users up and running on the system. They all have their logins, workspace, and the ability to submit jobs worked from day one.” says Fabrizio who is especially pleased with the pace of things “People in our group have already processed thousands of jobs and analysed a few Tb of ATLAS data.” Now that they are a full grid site, processing jobs for ATLAS the next step is to work on EPP’s other experiments SNO+ and nEDM.

The new system at Sussex is:

• 30 R410 nodes , each with 12 cores @ 2.66Ghz with 4GB/core
• 3 Interlagos nodes, each with 64 cores @ 2.4GHz with 2GB/core
• 2 older AMD nodes, each with 48 Cores @ 2.1GHz with 5.3GB/core
• 16 older intel nodes,each with 8 cores @2.4GHz with 1.5GB/core
• 162 Tb of disk space, using Lustre
• 1Gb connection to JANET

All researchers working on the grid will have particular software requirements and the systems administrators have to keep on top of these. One way the grid simplifies this is by grouing people working on similar experiments or science together into Virtual Organisations (VO), This helps simplify communication, requirement requests and software announcements, however it’s not a cure-all solution.

When you are looking after a computer cluster with a massive number of users and applications, automation is the name of the game. If a sys admin had to manually check every user, every data source and every application running on their machines they’d need more than 24 hours in a day. Unsurprisingly a lot of these processes are automated but when it comes to VO software it’s not the updating that’s a problem, it’s knowing that it needs to be updated. “The tools we use to configure the software at our sites is pretty good at keeping everything up to date” says Stephen Jones at the University of Liverpool “but it first has to know an update is available. This all hinges on the correct information being available to sys admins and them doing a manual update”. For GridPP sites this information is on a wiki page, a page which was updated by hand, adding extra unreliability.

dog nose by Mark Watson

However Stephen realised that the EGI Operations Portal could hold the key. The portal is a large database containing a wide variety of information about every VO on the grid, including what users are part of it, what resources they have access to and what software it uses. “All the information we need is there” explains Stephen “but it needs to be trawled through, converted manually and the sys admin has to actually remember to do it regularly. This makes the task labour intensive and unreliable. I knew we could do better than that”.

When Stephen started to investigate he discovered that the information was available in a flexible format that can be used to present data so both machines and humans can understand it. “This was important” says Stephen “if a machine can read it, then it can be automated”.

By June last year Stephen had developed the VomsSnooper toolkit so it could keep the wiki page up to date, check that a sites software configuration matched those in the Operations Portal and if it did not create the correct new files for the sys admin. “I was very happy with my progress” says Stephen “but once the application could ‘read’ the information on the Portal I realised I could do a lot more so started adding functionality and thinking of different use cases. Now VomsSnooper can actually perform 10 different tasks“.

This latest release includes updates to extend VomsSnooper to make introducing new VOs easier and simplifying a roll back if any problems happen with newly installed software. “As new, small VOs become common, we need a decent way to bring them into the grid quickly and with the minimum fuss. ” says Stephen “I hope VomsSnooper can make this a reality and would like to more sites adopting the toolkit”. Stephen will be presenting his work at the EGI Community forum, which is in Manchester from April 8th to 12th.

More information on deploying and using the VomsSnooper tools is available on the wiki page http://www.gridpp.ac.uk/wiki/VomsSnooper_Tools.

The kind of research being done at the Large Hadron Collider at CERN would be near impossible without the grid. This has intrigued many other particle physics experiments and they have been looking to get up and running on the grid. However moving from one or two people running test jobs to something that the entire community can use is not easy. “We’ve found ourselves in a bit of a difficult situation in that we’re a small experiment, but we need more computing resources than would be available at one university’s cluster.” says Matthew Mottram who works on SNO+ at the University of Sussex “Alongside this we have members across the world and a number of computing systems with different methods for submitting, running and monitoring computing jobs. Accessing the grid using the scripts developed by the SNO+ community during testing just wasn’t scalable.”

SNO+ are not alone, researchers from both T2K and SuperB are also looking to fit the grid into their workflow. All three look at what is called flavour physics and all have come to the conclusion that Ganga probably fits their needs. Developed for use by ATLAS and LHCb, Ganga is a simple application that is used to create, submit and monitor jobs, most importantly it can be used with just a single machine, a local cluster of computers or the grid. “Ganga allows for jobs to be submitted to a range of computing architectures requiring minimal configuration.” says Matthew “Their system also makes writing plugins to extend its use to other experiments very easy.”

This extendibility is one of Ganga’s major selling points. It was even chosen by a small company called Imense when running their image tagging work on the grid. “We always wanted Ganga to be customisable to meet the requirements of any new user communities, that’s why we implemented the plugin idea” explains Ulrik Egede head of the Ganga project “This means that once someone has investigated and built a solution for their community with Ganga it can be used by an end-user to process tasks without needing detailed knowledge of the grid.” It was actually SuperB that was first of the three to use Ganga. Armando Fella from the INFN facility in Pisa works on SuperB and has been instrumental in writing their plugin, “SuperB has not been built yet but we already realise the importance of having a reliable, accessible IT infrastructure in place before data taking and for simulated data analysis” he explains “last year we started looking into Ganga as our grid interface for distributed data analysis and simulation purposes. We had a small enough team but last May our code was included in Ganga so our users now have a simple to use interface.”

The SNO+ Detector

T2K is in a similar situation to SNO+ but has been using the grid for a bit longer and now wants to make it accessible to everyone in the collaboration “To help me understand Ganga I have been using scripts that can interact with the application and do the required tasks” says Luke Southwell from Lancaster University “now I just need to mould these into a basic plugin so anyone in T2K can run jobs. The SuperB plugin has been very useful and I’ve been working with the SNO+ people as well.”

Not only has Ganga proved very flexible for physics but smaller teams have used it. “I have worked with a few researchers interested in using the grid” says Mark Slater from the University of Birmingham “and I usually suggest that they use Ganga. It solves a lot of problems and makes life easier once the community reaches a critical mass.” The next experiment on Mark’s list is again particle physics with CERN-based experiment NA62 starting to investigate use of the grid.

If you are interested in using Ganga or the grid for your research you can find more technical information here or contact GridPP on info@gridpp.ac.uk.

The 2012 EGI Technical forum was the 3rd organised by the European Grid infrastructure to complement their community forum held earlier in the year. The event’s theme was The development of an open and sustainable EGI ecosystem that will support Open Science in the digital European Research Area. The UK was well represented by attendees from GridPP and other projects contributing sessions, demos and talks in the areas of security, accounting, tools, dissemination and much more. The Technical Forum is a big date in the grid calendar so is the perfect opportunity for announcements. This year was no different with the news that Globus Online is available for European researchers as well as the launch of the EGI Champions and a public demo of the EGI federated cloud. The meeting was also the start of the promotion of the next EGI Community Forum, which is in Manchester between the 8th and 12th of April 2013. Hosted by the UK NGI (of which GridPP is a member) it marks the first European grid meeting in the UK since Enabling Grids for EsciencE’s User Forum in 2007.

Before that though GridPP were back in the UK last week for their biannual collaboration meeting. GridPP29 was held in Oxford as a chance for the entire community to come together and have a full and frank discussion on the pressing topics that they are facing. The focus for the meeting was evolution, looking at where the service and its users are going and what needs to change.

The 2 days had talks from all the usual suspects like the sites, experiments and other stakeholders, as well as lively discussion and debate. This time though there was also presentations from the DiRAC HPC facility, the e-Infrastructure South Consortium and JANET ensuring that the meeting covered every possible area of interest to the community.

The GridPP Collaboration at GridPP29 in Oxford

All of the talks and material from the GridPP collaboration meeting are available at http://www.gridpp.ac.uk/gridpp29

You can find more information about the Technical Forum at http://tf2012.egi.eu/ with the full programme online at https://indico.egi.eu/indico/conferenceDisplay.py?confId=1019.

The announcement follows the news earlier in the summer that NGS had secured funding until July 2013. As NGS the project has developed a set of core services supporting the national grid infrastructure and now they want to extend these to other kinds of computing resources. “Changing the name to NES recognises that we must support all areas to create an integrated national e-infrastructure.” explains David Wallom, NES Technical Director “This includes cloud computing, HPC, big data and others. They all require a stable set of foundations on which to base user facing services, whilst recognising that different problems require different solutions.” “We were very happy to hear that NGS would continue to be funded by JISC and EPSRC last July” says Peter Gronbech, GridPP Project Manager “we work very closely with them as part of the UK National Grid Initiative for the European Grid Infrastructure. The name change reflects the broader landscape we both fit into.”

Roses by slgckgc

The NES is an important component of the UK e-science community and Neil Chue Hong from the Software Sustainability Institute has been working in the area for a number of years. He is excited to see what the future of NES holds “For 8 years the NGS has provided key infrastructure and services such as the UK Certification Authority to thousands of users. The change in name to e-Infrastructure recognises the underpinning role that the new NES plays in supporting researchers across the UK.”

More information about NES can be found at http://www.ngs.ac.uk/.

Held amid the chaos of the Edinburgh Fringe the Turing Festival is a 3 day celebration of the effect digital technology and the web has, and can have, on our world. Named after Alan Turing the father of modern computing, the festival aims to attract people from all walks of life encouraging them to discuss, collaborate and explore the ways in which technology is affecting all aspects of culture and society. It even had TuringTown a community of locals offering free accommodation to attendees, a real demonstration of the festival’s audience, people looking to help people.

The ethos of the festival closely mirrors the particle physics community and in particular CERN, the European particle physics laboratory on the Swiss/French border. The research being done at the lab would be impossible without cutting edge technology and a large cooperative community. Since the 1950s they have been consistently pushing the boundaries of engineering, computing and science. Public interest in the field has increased in recent years. With CERN being the birthplace of the worldwide web and evidence from the LHC for the Higgs boson mounting, 2012 seemed a great time to bring CERN, the LHC and grid to the Turing Festival. So earlier this year GridPP was approached to select speakers for the opening session at Turing Fest 2012.

The final line-up for CERN: Big Questions, Big Science, Big Technology included some of the biggest names in the community. It all kicked off with Professor Tejinder Singh Virdee from Imperial College London who led us through how the LHC and its experiments were conceived, including stories about visiting well guarded “secret” factories in Russia growing the crystals used by the CMS experiment. After Prof. Virdee came one of the world’s best known theoretical physicists Prof John Ellis with a pun-filled and engaging talk about physics before and after Higgs. With the physics explained it was the turn of Dr Ben Segal of CERN to talk about the history of computing at CERN and its evolution in an environment where it plays a secondary role to the research. Ben also worked in the same team as the creator of the web, Tim Berners-Lee and was able to give an illuminating view of his time at CERN. To wrap up the session was Prof David Britton, GridPP Project Leader, who discussed CERN’s “supercomputer” the grid, its history and its future. Just over a decade old the grid has now become an invaluable tool for the LHC and many other disciplines and David has been involved the whole time while still being working on the ATLAS and CMS experiments.

Alan Turing sculpture at Bletchley Park by Chris Brown

Each speaker was well received with many questions from the audience (and from the other speakers) and a short discussion at the end. There was one surprise in store for the attendees as during the final talk a very special guest made an appearance, Professor Peter Higgs. As one of the major theoreticians behind the eponymous boson it was a treat for the audience who greeted him with a well-earned round of applause.

“It seems appropriate that the 2012 Alan Turing centenary should be marked by such world changing announcements from CERN on the fundamental physics of the Universe. The CERN and GridPP stories exemplify the central role that computing has in our lives from aiding in scientific discovery to changes in art, culture and society.” says Dr Jamie Coleman, the director of the Turing Festival “It was an honour to have the leading lights behind CERN and the LHC experiments in Edinburgh, representing the heart and soul of the Turing Festival and to have Professor Higgs himself was the icing on the cake.”

Prof John Ellis, courtesy of CERN

The speakers and talks will be: Professor Tejinder Singh Virdee from Imperial College London Physics at the LHC, conceiving and constructing the accelerator and its experiments Dr Ben Segal of CERN – Computing at CERN from Wim Klein to Tim Berners-Lee and beyond. Professor John Ellis from Kings College London – Physics before and after the LHC and where will we go next? Professor David Britton, GridPP’s project leader – The Grid, CERN’s Supercomputer

Prof Jim Virdee, courtesy of CERN

Steve Wozniak, courtesy of Bryan Gosline

Outside the GridPP event one of the highlights of the festival is the keynote talk from Steve Wozniak, the entrepreneur and co-founder of Apple Inc on Thursday evening. The rest of the festival, however, promises to be just as exciting with sessions on citizen science, mashups, digital publishing alongside debates and workshops. With 3 very full days, and closing with a massive party in Summerhall, Turing Fest promises to be a fascinating and diverse event with something for everyone.

Tickets are moving fast for CERN: Big Questions, Big Science, Big Technology so get in soon to have a chance to hear these amazing talks and quiz the people who are involved in one of the greatest scientific discoveries of the last 50 years.

The details for the GridPP-organised session are:
Date: Thursday 23rd of August
Time: 09:00 to 12:00
Location: Theatre 4, Appleton Tower, Crichton Street, Edinburgh, EH8 9LE
Tickets: £15 in advance from Eventbrite
More information: http://www.turingfestival.com/cern/

The Higgs boson is the elementary particle needed by the Standard Model, particle physics’ “theory of everything”. The Standard Model works perfectly as long as the fundamental particles have no mass. Sadly not only do they have mass but also the mechanism that gives them mass was unknown. Then in the 1960s a number of physicists, including Peter Higgs, postulated a field that would permeate space, interacting with particles and giving them mass. In physics every field has a ‘carrier’ particle, called a boson. For the ‘Higgs’ field to exist it needed a ‘Higgs’ boson and so the hunt was on to find evidence of, what would turn out to be, a very elusive particle.

The Standard Model

Particle physics relies on two things; mathematics and particle accelerators. The first is where it all begins, attempting to explain how the world works through understanding the underlying maths. The second is the fun bit, researchers accelerate particles to high speeds and collide them with each other or against a target. The current state-of-the-art accelerator is the Large Hadron Collider (LHC), a 27km ring buried under the French/Swiss border accelerating protons to near the speed of light and colliding them. The LHC’s collisions recreate the conditions of the early universe, produce particles that we don’t see “in the wild” and help explain how the universe works. One of the particles it was hoped the LHC would create was the Higgs.

A candidate for a Higgs to Gamma Gamma decay in the CMS detector

“This could be the last piece of the standard model” explains Dr. David Colling, from Imperial College London who works on GridPP and Higgs searches, “It could also prove to be the first glimpse of something beyond it. We are definitely on the verge of something interesting and we have a fascinating few years to come”.

The LHC is based at the European Organization for Nuclear Research (CERN) and is not the first accelerator that hoped to produce the Higgs. Two of the most important were the Tevatron in the United States of America and LEP (the LHC’s predecessor at CERN), while neither proved the existence of the Higgs they were able to set a lower limit on its size and energy level, narrowing down the search a little. With these lower energies ruled out it was hoped that the LHC, at least 8 times more powerful than these machines, would be where the Higgs would be first seen. Alongside the massive engineering and energy advantage the LHC has over previous experiments, there is also the edge given by the grid. With over 200,000 computers connected worldwide, the grid provides 24/7 access to the computational resources needed to understand the data deluge coming from the experiment. This means that researchers can process large datasets and extract the meaning from the collisions (600 million every second) being produced at CERN.

Peter Higgs visiting the CMS detector in 2008, courtesy of CERN

“The performance of the GridPP as part of the worldwide LHC Computing Grid has been fantastic. We have been a leader in this area for the last ten years and I couldn’t be more proud of what we have helped achieve”, says Prof. David Britton, GridPP Project Leader. “The global community has allowed vast quantities of data to be analysed in an astonishingly short time. The last data that the ATLAS experiment included in their analyses had been taken less than a week earlier”.

This announcment is not the end of the story for either the LHC or the grid. There are many more exciting areas of research that use the world’s largest machine including studying the top quark, as talked about in the 3rd “Stories from the grid” episode. And the grid is now a major tool for researchers across the globe working on disciplines as diverse as the humanities and computational chemistry.

The top quark is one of the fundamental building blocks of the Universe and is one of the particles being studied at the LHC. It could help physicists understand why there are deviations from the standard model and potentially explain many mysteries like dark matter and extra dimensions.

The research is powered by grid computing, which is indispensable. In the video, Marcel Vreeswijk and Hurng-Chun Lee from NIKHEF, the Dutch National Institute for Subatomic Physics, explain how the grid is key to analysing and managing the massive sets of data. Without these tools, it would be impossible to pick out the key results that could hold the clues to top quark behaviour.

Previously “Stories from the grid” has highlighted structural biologists and engineers reviving lost instruments.

The video can be seen on YouTube: http://go.egi.eu/topquark