In 'uptime' output, what does "load average" mean, and when does the number become "bad"?

Posted on 2004-04-07
Medium Priority
Last Modified: 2011-10-03
According to the AIX man page for the 'uptime' command, the following is true:

"The uptime command prints the current time, the length of time the system has been up, the number of users online, and the load average. The load average is the number of runnable processes over the preceding 5-, 10-, 15-minute intervals."

My numbers are usually below one, so I haven't been concerned.  Lately, the numbers are sometimes getting higher.

Can you tell me what the "load average" or "runnable processes" means, and at what point the number becomes "bad"?

In other words, is 1.5 okay, but 5.3 bad?  The man page is not very informative on this issue.

Thank you!!

Unca Xitron
Question by:Xitron
LVL 48

Accepted Solution

Tintin earned 500 total points
ID: 10779306
The load average is the average number of processes in the run queue.  By that, it means the number of processes waiting to be processed by the CPU.  The uptime command gives you a 5,10 and 15 minute average so that you can see if the server performance  is getting worse/better/stable.

As to what a good and bad figure is, depends a bit on the system and type of processes running.  As a very rough rule of thumb, when the load average gets above 10, you may experience some minor performance issues.  However, I've been on some systems where we've had the load average over 500 and the system coped OK (obviously not as efficient as when it is in normal operation).

Load average is just one component at looking at general performance.  You really need to look at it in conjunction with memory and I/O to get an overall view.
LVL 45

Assisted Solution

sunnycoder earned 500 total points
ID: 10781025
>Can you tell me what the "load average" or "runnable processes" means, and at what point the
>number becomes "bad"?
There are 5 main stages in the lifetime of a process
1. Entry .... when it is just invoked
2. running ... when it has a CPU and is running
3. runnable ... when it is ready to run and is waiting for a CPU
4. waiting .... when it is blocked waiting for something such as I/O or semaphore
5. exited/stopped

from OS point of view, we often consider 2+3 together as runnable and divide 5 to incorporate zombies ...

So for your question a runnable process is any process which is ready to execute ... it may or may not have CPU at the current moment ....

OS maintains a queue of the runnable processes ... Whenever currently executing task has completed or it needs to be rescheduled or blocked, OS just picks a process from the runqueue and starts executing it ....

So load average would mean that over last n minutes, at an average, how many tasks were in the runqueue ....

I hope it makes some sense
LVL 62

Expert Comment

ID: 10782987
Depends on how you tune system (task scheduler, swap management, disk buffers, mealloc() )
Install and patch bos.adt.samples package from 1st CD, so you get vmtune utiliy to tune malloc parameters, especially swap threshold can be multiple megabytes instead of default one, and read the README as dmesg said, to tune individual applications.
Check out "smitty wlm" , it has very simple interface and helps you group and prioritize applications as necessary ( first try monitoring mode to see how well you grouped your processes )
Kernel min=20% max=100% cpu and memory (root is included here, do not lower that 20%)
Interactive logins (gid=users) -"-
Database application (uid=postmaster) -"-
Webserver (program=/usr/IBMHTTPServer/bin/httpd)-"-
default -"-
i.e. you do not know how much resource each one needs, so give them equal share of resources, so they all run at once without bringing others down, then you will gain experience who needs a bit more or less and adjust settings accordingly.
I run 12 workload classes on one test server, this becomes less on actual business servers

Author Comment

ID: 10784069
My thanks to all for your excellent information!

Unca Xitron

Featured Post

Concerto Cloud for Software Providers & ISVs

Can Concerto Cloud Services help you focus on evolving your application offerings, while delivering the best cloud experience to your customers? From DevOps to revenue models and customer support, the answer is yes!

Learn how Concerto can help you.

Question has a verified solution.

If you are experiencing a similar issue, please ask a related question

A metadevice consists of one or more devices (slices). It can be expanded by adding slices. Then, it can be grown to fill a larger space while the file system is in use. However, not all UNIX file systems (UFS) can be expanded this way. The conca…
FreeBSD on EC2 FreeBSD (https://www.freebsd.org) is a robust Unix-like operating system that has been around for many years. FreeBSD is available on Amazon EC2 through Amazon Machine Images (AMIs) provided by FreeBSD developer and security office…
Learn several ways to interact with files and get file information from the bash shell. ls lists the contents of a directory: Using the -a flag displays hidden files: Using the -l flag formats the output in a long list: The file command gives us mor…
In a previous video, we went over how to export a DynamoDB table into Amazon S3.  In this video, we show how to load the export from S3 into a DynamoDB table.
Suggested Courses
Course of the Month15 days, 14 hours left to enroll

850 members asked questions and received personalized solutions in the past 7 days.

Join the community of 500,000 technology professionals and ask your questions.

Join & Ask a Question