Operating System - Tru64 Unix
cancel
Showing results for 
Search instead for 
Did you mean: 

tempreture problem for alpha server 4100

MahdiRa
Frequent Advisor

tempreture problem for alpha server 4100

Hi all
i have an alpha server 4100 installed in room with very good cooling system the new problem that appears since 5 days if i just opend the door for a while the system is crashed and the tempreture reached 51degreeC which the maximum that the system crash after. i do clean all the fans of the CPU and power supply but the system retun to the same problem i used the show power command from the console and the result is good. plz advise
20 REPLIES
Mobeen_1
Esteemed Contributor

Re: tempreture problem for alpha server 4100

Mahdi,
My opinion is that there is not much you can do other than make sure that you have the right fan.

I would like you to explore the idea of adding an additional fan and may be at the same time position this fan in such a way that the the air is sucked out instead of blowing.

Try this, all the very best :)

regards
Mobeen
MahdiRa
Frequent Advisor

Re: tempreture problem for alpha server 4100

i agree with you that this is good but what i need is how to trubleshoot and to solve the problem. i thinking to change all the fans inside the server but i couldnot fined any thing in the loge files that said the problem in cooling or any thing else
Michael Schulte zur Sur
Honored Contributor

Re: tempreture problem for alpha server 4100

Hi,

about which door are you talking, the front door?

greetings,

Michael

MahdiRa
Frequent Advisor

Re: tempreture problem for alpha server 4100

am not talking about the door but any change in the room condition has the same effect in crashing the system
Mobeen_1
Esteemed Contributor

Re: tempreture problem for alpha server 4100

Can you run up the SDA (stack dump analyzer) and issue some commands, the details of which are as follows

See one of the things is, when your system is already down to the console and ready to boot up, the power should be good, else it would never have booted if the temp was high. I thought there was a way to look at the history of temp while at console. Never the less, you should be able to see all that you need using the SDA.

$:/>ANALYSE/SYSTEM
SDA:/>SHOW CRASH
SDA:/>SHOW SUMMARY
SDA:/>SHOW CPU

These typically give you the values at the time of crash.

Let me know how it goes

regards
Mobeen
Mobeen_1
Esteemed Contributor

Re: tempreture problem for alpha server 4100

Mahdi,
I am sorry i thought you were running OpenVMS, hence suggested to run ANALYSE/SYSTEM. Do you know how you invoke SDA (dump analyser) in Tru64. If you know do that and look at the following

SDA:/>SHOW CRASH
SDA:/>SHOW SUMMARY
SDA:/>SHOW CPU

regards
Mobeen
Michael Schulte zur Sur
Honored Contributor

Re: tempreture problem for alpha server 4100

Remove the fron door of the 4100.
It will get better air.

greetings,

Michael
MahdiRa
Frequent Advisor

Re: tempreture problem for alpha server 4100

no i dont know how to do it so if you can help me or any one else to do it
Mobeen_1
Esteemed Contributor

Re: tempreture problem for alpha server 4100

Michael,
Do you know what you need to run on Tru64 to run the equivalent of Analyse/system (for SDA) in OpenVMS

regards
Mobeen
Mobeen_1
Esteemed Contributor

Re: tempreture problem for alpha server 4100

Mahdi,
Look through the link below if it throws any light

http://h30097.www3.hp.com/docs/base_doc/DOCUMENTATION/V51B_HTML/MAN/MAN5/0090____.HTM

regards
Mobeen
Mohamed K Ahmed
Trusted Contributor

Re: tempreture problem for alpha server 4100

Mahdi,
I have an old 4100 system here, and what I did to get around this problem is that I have removed the back door of the system cabinet.

I think it is a good idea to remove both front and back door of the cabinet and monitor.

Mohamed
Mohamed K Ahmed
Trusted Contributor

Re: tempreture problem for alpha server 4100

Mahdi,
Have you tried connecting to the system through the RMC console and checking the temp while the system is up and see the change

Mohamed
MahdiRa
Frequent Advisor

Re: tempreture problem for alpha server 4100

but it was running very good befor with very bad condition related to the current situation now so why to remove any thing there is somthing rong it must be fixed
Mohamed K Ahmed
Trusted Contributor

Re: tempreture problem for alpha server 4100

May be the thermal sensor is not working
Log a call with HP support and let them test the sensor.

Mohamed
MahdiRa
Frequent Advisor

Re: tempreture problem for alpha server 4100

so do any one know how to test this sensore
Michael Schulte zur Sur
Honored Contributor

Re: tempreture problem for alpha server 4100

Hi,

do you have decevent and can look for errors?
There are a few things on console level, you can look like show power ..
Please look with help.
Look if there is a crash file.
Look into the crash data for the panic string.

You should contact HP, if you don't have a fan failure or blockage.

Did the machine report anything while booting/resetting?

Michael
MahdiRa
Frequent Advisor

Re: tempreture problem for alpha server 4100

no every thing is usual nothing strange and am not sure but nearly i can agree that th fans are running in good way
Mohamed K Ahmed
Trusted Contributor

Re: tempreture problem for alpha server 4100

Call HP support and let them come in and check the sensors.
In the mean time, just put a fan and direct ot to blow underneath the system

Mohamed
Hein van den Heuvel
Honored Contributor

Re: tempreture problem for alpha server 4100


It should be easy to verify whether the system is truly how, versus it thinking that it is dangerously hot. Just stick a few thermometers, prefereable min/max types aroudn the box no?

In the mean time, your description so far suggests to me that it really is too hot.
So you have an environmental problem.

It this system rack-mount?
Is is venting the same direction (front to back!?) as the other systems in the rack, notably those above an below it?

If it is venting in the opposite direction then it will just suck in hot air blown by the system next to it, heat up some more, exhaust, be sucked back into neight, which heats some more still... and so on untill the alarm goes off.

So check those airflows! (with a sheet of paper before the fans ?).
Temporarely use an external fan to break the loop with fresh air.
Potentialy install an air baffle to divert/break the air-short-circuit.

Are the ES40 fans (re)installed backwards (ok, I do not know whether that is even possible).

Hope this helps some,
Hein.

Michael Schulte zur Sur
Honored Contributor

Re: tempreture problem for alpha server 4100

If the environment monitoring is activated, you can monitor the temparature with:
sysconfig -q envmon

Michael