>>> On 9/27/2011 at 01:06 PM, "Therrian, James" <[email protected]> wrote: > There is application being developed here running Message Broker on z/LINUX. > During many of the tests the distributed guys are focusing on "SYSTEM LOAD" > and it's value during our tests. While the tests seem to be running > acceptable they are still concerned about the system load value. Can someone > help in explaining what metrics go into the SYSTEM LOAD value and the how to > interpret differences of that value between zlinux and linux/unix if there > are any?
Essentially, system load is calculated by taking the average of the number of processes in states R and D over the last 1, 5, and 15 minutes. It should not be any different across architectures. Focusing on system load having a particular value is really meaningless, since what it "should' be varies across different workload types. A system load of 40 might be perfectly acceptable for one type, and absolutely horrible for another. Essentially, you need to figure out what the system load looks like when the workload is running well, and compare future observations relative to that. Too low of a system load can be just as troublesome as one that's too high. Mark Post ---------------------------------------------------------------------- For LINUX-390 subscribe / signoff / archive access instructions, send email to [email protected] with the message: INFO LINUX-390 or visit http://www.marist.edu/htbin/wlvindex?LINUX-390 ---------------------------------------------------------------------- For more information on Linux on System z, visit http://wiki.linuxvm.org/
