A database server is generally a mission critical component. Having a recovery plan is essential to ensure the business can continue to function in case of hardware failure, natural disaster, or other catastrophe.
A backup server or a standby server (using Database Mirroring, AlwaysOn Failover Cluster Instances, AlwaysOn Failover Groups or Log Shipping technologies) is an option, but not always practical or possible due to the costs of hardware and associated software. Without some kind of a standby, the key to recovery from a failure is to get the database server back up and running as soon as possible with the least possible loss of data.
Checkout my article here, in which I discussed about recovery procedures for a catastrophic failure. I also talked about how to view and manage database its file states. Finally, I talked about how to troubleshoot and resolve disk space issues with tempdb, and how to move tempdb.
This article is published on SSWUG.org.