UI cache corruption occurs during machine status assignment, causing the cache to become corrupted and messing with the visible machines list until the UI cache is deleted and regenerated.
This is caused by a UI code bug that is pulling the incorrect number of admin status columns from the SQL database, causing an extra 'empty' column to appear in the UI. When this column appears and is selected to clear the machines administrative status(which is typically why its chosen) it sends a broken UPDATE statement to SQL which then corrupts the UI's cache, until the cache is manually deleted.
Deleting the UI cache will resolve this issue until the next crash is encountered. The immediate workaround is awareness: do not select the blank field seen in the Machine Admin status column. This should allow for normal work to take place, always select a status that has a named description in the drop down box: