Sun Hardware Diag Tools

Applications Filesystems, Disks and Memory
1

Problem: Several possible cooked Sun U10's.

Need: Sun Hardware Diag Tools.

I want to pound on these boxes and see whats what. I am wondering where I could find some Diag tools that will help me diag problems. Freeware is preferred. (The magic budget of $0)

2

Solaris comes with some tools that you can use and others can be found on the web. Look into sar, vmstat, netstat, iostat, top (freeware), proctool (freeware).

Example - Sunperf.com

3

I had seen these performance tools, but are they going to let me stress test the hardware?
I want to load it up, so right now I have a CPU intensive application running.

4

If you are looking to put a load on the system, this can be accomplished with a simple script.

CPU Intensive:

#!/bin/ksh
while true
do
counter=$counter+1
done

Filesystem intensive:

# find ./ -name "*" > /dev/null

or

# mkfile 10m myfilename

(The last one should be done on a filesystem that can handle it - suggest a bigger file but that is something you can look into)

Someone else can give you something for memory - I've gotten it to climb before but not as much as with cpu and filesystems - and I can't seem to find the script I used for that or remember how I did it.

5

Just edit a file with emacs...that should stress both memory and cpu 8)

Actually, memory always has something in it. Consuming memory is an OS concept. And modern CPU's seldom halt. An "idle cpu" is another OS concept. Idle CPU's are scanning the run queues looking for something to do, and that idle loop is arguably "running a program". Are you trying to stress the system or locate hardware problems?

If you really want diagnostics, see: Hardware Diagnostics for Sun Systems: A Toolkit for System Administrators

6

I am trying to locate hardware problems that I know exists in one form or another on these three boxes.

7

I don't know if I am way off for what you are asking, but go to the "ok" prompt by typing "init 0". Once there type "setenv diag-switch? true" and "setenv diag-level? max" and "reset-all". It will spew all kinds of hardware info. This is just a basic overview, I suggest you go to docs.sun.com and read more about these commands if this is what you are looking for.

8

I've used the enbedded diags and get zero for problems. The info on the eeprom diags helped thanks!

The thing I am lacking is a solid answer to why these boxes are crashing. I think its a heat related problem, but shoot I have no clue how to prove it (hence why I was looking for a tool suite of some sorts). I check Sun site for any other diag methods.

Thanks all!