HUBzero + Cloud: P Power tool for the m masses un, Christopher Thompson n, Alex Yountss, Preston Sm mith and Caroll Song Dave Brau osen Center fo or Advanced Computing Ro Purd due Universityy Today’s powerful cybe erinfrastructure, including the high perfformance and d distributed computing systems aand clouds, caan be an unfo orgiving place for the non p power users. Users are oftten expected to be familiar w with Unix com mmand lines, w whether runn ning computaation jobs or m moving data ffrom one placce to another. W When they ne eed immediate help, userss typically relyy on emailingg technical support and sysstem administrators, while in the commeercial service provider worrld, online interactive help is fairly comm mon. we bridge thesse gaps and h help scientificc users to be m more producttive? We have developed How do w Springboa ard, a web plaatform based on the succeessful HUBzerro software, tto provide eassy access to H HPC, cloud and d other digitall resources an nd online collaboration capability for in nstant help. Springboard has been integrated with tthe TeraGrid, including acccess to all thee TeraGrid ressources and relevant informatio on (e.g., usagge, allocation,, etc). Leveragging the HUBzero middlew ware, especially the use of VNC (Virtual Network Comp puting), Sprin ngboard userss are able to sshare their on nline sessionss, including running o of tools and ap pplications, as well as login n and commaand windows to the compu uting resourcces, This capab bility not onlyy allows real‐ttime collaborration amongg the collaborators, but alsso provids an online interactive platfform that sup pport staff can n use to help their users. e a Springboaard GUI appliccation, Cloud Dashboard, tthat allows ussers to accesss This posteer will feature cloud reso ources from SSpringboard in their web b browsers thro ough a GUI. The Cloud Dasshboard connects from the SSpringboard w web portal to o Wispy, a clo oud resource provided by P Purdue Univeersity and a resource partner on th he Science Clo oud (http://w www.scienceclouds.org). W Wispy is a cluster running the d software. Itt accepts userr supplied virttual machine images and rruns each VM M KVM and Nimbus cloud with up to o 4 CPUs and 16GB of mem mory. Users aare able to package their applications and operating systems ccompletely insside the imagges they subm mit and have ffull control ovver the execu ution environment. Wispy haas been used to provide sm mall, instant, on‐demand cclusters for various tasks aand running co omplicated o or prepackageed application ns on addition nal hardware resources. Th here are currently 32 physicaal hosts (128 cores) on Wispy with a capacity limit o of 128 VMs. Cloud Dasshboard supp ports both battch and interactive, or on‐demand, types off applicationss. From its G GUI, a user caan load multip ple VM imagees and run th hem on Wispy. The user m may choose th he length of runtime of th he VM, wheth her to save th he image upo on terminatio on, and with o one click, can n launch thee VMs in the cloud. The VM M status is reported back to the u user in the GU UI. While the user may connect to the VMs VMs are rrunning, the u via one bu utton click an nd a separate login window w is then displayed (Figure 2). The user may also choose to o issue a comm mand to be eexecuted on aa VM via the Dashboard (Figure 3). In ntegration witth additional cloud resources is in the plan. Figure e 1. Illustration of Springboarrd, Wispy and Nimbu us Tools such as the Cloud Dashboard have the potential to change the way HPC, cloud and other digital resources are typically used today, i.e., via command lines, which means a steep learning curve for a lot of people. These tools will enhance the usability of the HPC and cloud resources and allows the researchers to focus their attention and effort on scientific discovery rather than the intricate details of the computing and software environments. Together with the HUBzero’s community participation and collaboration tools, Springboard hopes to bring the high end computing and data resources to the masses! Figure 2. Cloud Dashboard GUI and VM access window Figure 3.Running a command in VM via Cloud Dashboard GUI