Thoughts on Virtualization - You will soon have a chance to read our thoughts on the topic of Virtualization. In the meantime, insideHPC magazine is running an interesting series on this subject. Click here to read more.
X-ISS Webinar Available Online - If you missed our free webinar, “Cognitive Analytics for HPC Data Centers,” you can still watch the recording online by clicking here. Thanks again to the Society for HPC Professionals, which hosted the webinar.
Visit X-ISS at SC14 in New Orleans
Be sure to visit us in November at Super Computing 2014 where our theme will be, “Manage Your HPC to Maximize Your ROI.” X-ISS CEO Deepak Khosla and the X-ISS team will be on hand to discuss how your cluster can be designed, deployed, and configured to run at peak efficiency. You may be surprised to learn that our management services and tools consistently outperform internal staff. Let us show you how we put our 20 years of HPC management, deployment and troubleshooting experience to work for every client to maximize system performance and minimize downtime. Stop by booth #3760 to learn more!
Visit http://sc14.supercomputing.org/ for more info on the conference.
xCAT Upgrade with No Downtime...and BitTorrent to Boot
X-ISS successfully upgraded a client's cluster manager from xCAT 2.3 to xCAT 2.8. The large gap in software versions was the first challenge, but the client also wanted the project completed without any system downtime. In addition, we implemented BitTorrent as the compute node image transfer technology to provide the client with faster boot times using fewer service nodes. Going into the project, we found no documentation that BitTorrent had ever been deployed on an xCAT cluster.
Did we set our sights too high on this project? Read here to find out.
Personalized Service Helps BRP Keep the Engines Running
At X-ISS, we take great pride in our decades of expertise in managing HPC clusters for clients. For at least one of those clients – a developer of motorboat engines – our personalized service makes a difference, too. As part of our ManagedHPC® contract with the client, we took charge of an upgrade to their cluster, which involved the addition of extra nodes. When things got complicated, our insight into their system helped us smooth over the problem before more serious issues arose. In the end, the client credited this personal relationship for the quick resolution to the situation.
Read the whole story here.
HPC Data Center Relocated with 99.97% Uptime
A major X-ISS client put us in charge of managing the upgrade and relocation of its HPC data center. The move involved over 300 racks containing more than 7,000 nodes, networking equipment and 10 petabytes of storage. Downtime was something the busy client couldn’t afford. The six-month deadline didn’t make things easier. But the X-ISS team pulled off the move without a hitch by designing and implementing an extension of the network between the old and new data centers utilizing Arista network switches. The result was a move with a remarkable uptime of 99.97%.
Read how we did it here.