Load average shows how many processes are waiting in the run queue for a system resource (usually a processor). The higher the load average, the more processes are waiting.
One way to determine whether a machine has high load average is to use an operating system command such as
top while the application is running.
uptime output below shows that the load average is 0.40, 0.46, 0.43 over the last 1, 5 and 15 minutes, respectively.
top output shows, among other things, load average, CPU usage percentages and I/O wait (iowait) percentage. The iowait percentage is the percentage of time the CPU is waiting for an I/O to complete. The output below shows a fairly high load average over the past 1 minute (10.40) for the number of CPUs. It also shows that the CPUs are mostly in use (idle=3.0%) and that I/O wait percentage is low (0.4%).
Another way to determine whether a machine has high load average is to use
vsd to display the load average values contained in a given Geode statistics archive.
The chart below shows the LinuxSystemStats loadAverage1, loadAverage5 andloadAverage15 values.
Determining that there is high load is one thing. Finding the source of the load is another (whether it be CPU or I/O). Use the output from
top to help determine if the high load is due to CPU or I/O. If the high load is due to CPU, see the Action section of theTroubleshooting CPU document for actions that can help alleviate CPU issues. If the high load is due to I/O wait percentage, then it might be related to disk I/O. See the Apache Geode Documentation for actions that can help alleviate disk issues for persistent regions.