I recently moved the tempdb to a different drive and in the process totally removed 3 extra ndf files that were no longer needed. Tempdb currently resides on the f:\. After moving the tempdb and restarting sql to verify it worked we removed the drives U:\, g:\ and J:\ from the server. The old place where the ndf files for the tempdb were located. As you can see from the sys.master_files the ndf files are still being referenced. This is a 2 node cluster and the issue is when it fails over to node 2, sql won't start saying that it can find the extra ndf files. When I run sys.database_files on the tempdb it verifies that there is only 1 data and 1 log file. Is there anyway to remove the entries from the sys.master_files as this is causing our cluster to be useless?