RCC to move systems to NWRDC

Posted on

As the RCC grows, it is important for us to provide higher level of reliabilty to our research partners.  To this end, we are consolidating all RCC equipment to the Northwest Regional Datacenter (NWRDC).  RCC staff will remain in Dirac Science Library.

NOTE: This plan was subsequently cancelled for a number of technical reasons. It was subsequently decided to consolidate all RCC resources in the Sliger Data Center.  Details are available in a more recent news article.

As part of the University's ongoing effort to improve and consolidate data center operations, the Research Computing Center is moving all equipment to the Northwest Regional Datacenter. We do not have a definite timetable for this move, but it will almost certainly occur within the next 12 months during 2017.

This project will involve moving all RCC systems, including the HPC, Spear, Lustre, Panasas, Virtual Machines, and all support systems to the NWRDC. We are going to execute this extensive operation in a series of phases, which we are currently planning. Our objective is to generate as little service disruption as possible. As of now, the tentative plan (subject to change) is to move systems in the following order:

  1. Lustre, NoleStor and Management systems from our Dirac location
  2. Compute equipement from our Dirac location; this includes some HPC nodes, Spear nodes, and "SKY" Virtualization Cluster nodes
  3. HPC nodes and management systems from the Sliger location

Our move to NWRDC will benefit RCC users in many ways.  Some of these include:

  • We will consolidate two datacenters into a single datacenter. Currently, approximately half of RCC systems are in the Sliger Datacenter, and the other half are in Dirac Science Library. The new configuration will simplify server management and networking configuration, and will decrease potential for system failure.
  • The NWRDC is a well-supported and well-staffed facility with many system redundancies and disaster-prevention provisions. This move will drastically reduce the risk for service disruptions caused by infrastructure dependencies; for example, the prolonged downtime after Hurricane Hermine damaged our Dirac facility.
  • Partnering with NWRDC will allow the RCC staff to offload responsibility for infrastructure (power, physical space management, etc) to a qualified team of professionals. This will free up staff time to focus on managing higher-level tasks.

We will communicate more about this plan as it continues to evolve.  In the meantime, please let us know if you have any questions or concerns.