Hi Guys,
Tomorrow I am taking a RH exam and I am doing some practical questions with different scenarios. I have a question about the below question:

There is a client with high load average issue.
Troubleshoot this issue to find out what is responsible for this condition. What utilities would you use? Please document your thought process as well as what you would look for in the output of the commands you would run.

So far I summarize the below items:

a. Run “uptime” command and check what is the load average
b. Run “top” command an check CPU bound Load, Memory Bound Load and IO bound load… check for us,sy,id,wa.
c. Then check for Ram using the line “mem” and “swap” in “top” command. Then check for swap usage if swap usage is more, then it can lead to more IO.
d. Then check for IO usage using iostat and iotop command. Usinf iostat command find which partition is used more and based on that drill down to which process is using that partition.

Am I on the right path? is there anything else to add?

You're pretty much on the right track.

Check /proc/cpuinfo to determine the number of cores a machine has, because seeing a load average of say 1.0 for a quad core machine means you are at 25% load, not 100. I use free to check memory and use the -t option to get totals. Uptime is good to know how long it all has been running, yes it also gives you the load averages, top will also give you the load averages.

The bottom lines are CPU time and memory are the key resources. If their system is slow, they may be out of memory, out of CPU time, or both. It may be because of their overall use case rather than one particular problem, but if the problem is defined stating that this is unusual, then these tools are correct directions.

Don't forget collectl and sysstat.