There were Internet backbone issues last night followed by problems at Rackspace, where the InWorldz servers are hosted. Our Apache Cassandra inventory clients are retrying operations, but on some regions, the problems seem to require a region restart to resolve. See this thread
As a result, and to avoid any question about some regions continuing to be broken, we're going to do grid-wide rolling restarts of all regions to correct any lingering inventory access issues, starting in a few minutes
. These restarts will result in each region being shut down and restarted and each region is expected to be down for 15-30 minutes at various points during the evening.