|From:||Glenn Bresnahan, IS&T/Executive Director Research Computing Services (firstname.lastname@example.org)|
|Subject:||MGHPCC/SCC Outage August 11, 2014|
|Date:||July 25, 2014|
Each year the MGHPCC has a scheduled full outage for facility maintenance. This year the maintenance outage is scheduled for Monday, August 11. We will also be using this maintenance window to perform required maintenance on BU's Shared Computing Cluster (SCC) and our networking equipment. We expect the outage to last approximately 24 hours.
In future years, the annual outage will be scheduled in late May. This off-cycle maintenance is an intermediate service between last year's emergency service in December and the to-be regularly scheduled service in May.
In preparation for the outage, we will bring down the SCC cluster at 10:00 PM on Sunday, August 10. Starting now we will begin the process of draining the SCC batch queues, so that jobs which would not complete prior to the shutdown will remain pending until the system is returned to normal operation. During the maintenance outage MGHPCC will be servicing various electrical and mechanical components of the data center and we will be performing maintenance on various SCC systems and networking equipment. During this period we may be able to provide intermittent access to the SCC login nodes, but as the file systems are likely to be unavailable, the login nodes will have limited usefulness.
We expect operations to return to normal by midnight on Monday, August 11. However, we caution that a full shutdown increases the probability of unexpected complications which could extend the duration, as did happen in the December outage. As usual, we will post updates to the SCC status page at www.bu.edu/tech/research/sccupdates/.
Please email email@example.com or me directly if you have any additional questions or concerns.