Update: The rolling updates on compute nodes are in progress. The goal is to minimize disruption to job scheduling but that has turned out to be difficult due to the nature of our nodes having multiple queue instances. There turned out to be a high level of contention with high throughput jobs which caused those to fail. To mitigate that, the investor queue instances will be disabled on nodes that have not been updated yet. This will lead to jobs sitting in qw longer than usual. Please be patient as the compute get updated.
The Neon system will need to be patched and rebooted to mitigate a security that was announced yesterday. The login nodes will be rebooted today, February 17, 2016 at 2:00 PM. The exact time will posted when determined. Compute nodes will get rolling patches and reboots to minimize the impact on jobs.