eXcellence in IS Solutions (X-ISS), a provider of high-performance computing and big data solutions, has released Version 14.1 of its DecisionHPC analytics software for HPC cluster management. Upgrades include more comprehensive cluster scheduler reports and an innovative Attribute Heat Map visualisation.
DecisionHPC is a platform-neutral SaaS solution. In addition to monitoring technical performance of clusters, schedulers, and other resources, DecisionHPC adds business analytics related to projects, users, and applications.
A core capability of DecisionHPC is its ability to synthesise data from disparate hardware and software sources – including monitoring tools, resource managers, licence managers and schedulers. With this in mind, X-ISS has added SGE/UGE scheduler integration and made overall improvements to scheduler reporting. Version 14.1 offers a global view of queue usage and resource status as well as reporting capabilities that show how the configuration of cluster components varies across large geographically dispersed heterogeneous cluster environments.
‘This previously unavailable information gives management new data on how well jobs are running on the cluster to ensure that projects have been matched with the right resource and pinpoints where productivity can be improved,’ said Deepak Khosla, X-ISS President.
The dashboard visualises information on system operations from the macro to the micro level. Colour-coded Heat Maps have been expanded in the new release to represent both technical metrics as well as business attributes relating to configurations, users, projects and applications.
The Heat Maps show a graphical representation of cluster data where the individual values contained in a matrix are represented as colours showing which users, projects, and applications are processing and where they are running on the system. A homogeneous map means everything is functioning normally, but a variance on a single node is immediately recognisable by showing up in a different colour allowing technical or managerial staff to take immediate action.
DecisionHPC enables people within management to understand how system capacity is being used by specific users and applications. In addition, business analytics assist organisations in modelling how new applications will impact cluster performance, thus providing insight to make well informed decisions on HPC capacity planning and expansion.
‘Today more than ever, HPC Directors and technical staff need to make real-time business decisions based on data. DecisionHPC provides that additional insight with business analytics you can’t get anywhere else,’ said Khosla. ‘This solution enables HPC operations teams to understand how cluster performance is impacting the bottom line of the organisation.’