Redefining Load Average
Sena Seneviratne and David Levy have been cooking up some changes to the load average calculation in Linux. Their idea is that instead of just calculating a single number representing the entire load average on the system, multiple statistics should be available. For one, they want to separate disk I/O load from the rest of the load average calculation so it can be viewed independently. Also, they want to enable per-user load average calculation, so each user could see their effect on the system. Arjan van de Ven exhorted them to add statistics only and not to change the behavior of any numbers that the system already produces.
Comments