This paper analyzes the problem of checking the integrity of files stored on remote servers. Since servers are prone to successful attacks by malicious hackers, the result of simple integrity checks run on the servers cannot be trusted. Conversely, downloading the files from the server to the verifying host is impractical. Two solutions are proposed, based on challenge-response protocols.

No abstract

This paper presents a measurement-based dependability study using event logs collected during about 3 years from 133 Windows NT and 2K workstations and servers interconnected through a LAN. We focus on the identification of machine reboots, the classification of their causes, and the evaluation of statistics characterizing the uptimes, downtimes, and the availability of the Windows NT and 2K machines.

No abstract