[sepia] lab status, 11 October

Overnight, the iscsi service became unresponsive again, which caused a lot of services to go down/become unresponsive. I was able to manually restore the service and the VMs should be back and running as normal. I'm still investigating changes to the iscsi configuration to allow failover to happen more-gracefully (perhaps at all) that may make such temporary service failures easier to handle.