The network infrastructure affected some instance's ability to stay online. We are rebooting instances as required to bring them back online.
