[gentoo-user] Monitoring temperatures

[gentoo-user] Monitoring temperatures

[Grant]

I have another dead power supply and/or another dead motherboard in my
Gentoo router.  I've tried to make that system as silent as possible
and I wonder if I'm paying the price.  How do you guys monitor system
temperatures?  Is lm_sensors the way to go?  How do you keep an eye on
the temperatures of multiple local and remote systems?

- Grant

Re: [gentoo-user] Monitoring temperatures

[Paul Hartman]

On Fri, Mar 20, 2009 at 4:14 PM, Grant <emailgrant@gmail.com> wrote:
> I have another dead power supply and/or another dead motherboard in my
> Gentoo router.  I've tried to make that system as silent as possible
> and I wonder if I'm paying the price.  How do you guys monitor system
> temperatures?  Is lm_sensors the way to go?  How do you keep an eye on
> the temperatures of multiple local and remote systems?

As long as lm_sensors supports your sensor chipset you should have
plenty of options. Check the list of applications on
http://www.lm-sensors.org/wiki/UsefulLinks

I don't do any remote monitoring but I suppose you could use also use
lm_sensors + grep to do your own alert script. The basic 'sensors'
output on my machine looks like this:

abituguru3-isa-00e0
Adapter: ISA adapter
CPU Core:               +1.44 V  (min  +0.00 V, max  +1.65 V)
DDR2:                   +2.10 V  (min  +1.70 V, max  +2.50 V)
DDR2 VTT:               +1.05 V  (min  +0.85 V, max  +1.25 V)
CPU VTT:                +2.42 V  (min  +1.90 V, max  +2.90 V)
NB 1.2V:                +1.47 V  (min  +1.15 V, max  +1.75 V)
SB 1.5V:                +1.59 V  (min  +1.25 V, max  +1.85 V)
HyperTransport:         +1.27 V  (min  +1.00 V, max  +1.50 V)
ATX +12V (24-Pin):     +12.42 V  (min  +9.60 V, max +14.40 V)
ATX +12V (4-pin):      +12.54 V  (min  +9.60 V, max +14.40 V)
ATX +5V:                +5.04 V  (min  +3.99 V, max  +6.00 V)
ATX +3.3V:              +3.38 V  (min  +2.64 V, max  +3.94 V)
ATX 5VSB:               +5.22 V  (min  +3.99 V, max  +6.00 V)
CPU:                      +25°C  (high =   +75°C, crit =   +85°C)
System:                   +35°C  (high =   +55°C, crit =   +65°C)
PWM Phase1:               +56°C  (high =  +125°C, crit =  +135°C)
PWM Phase2:               +56°C  (high =  +125°C, crit =  +135°C)
PWM Phase3:               +56°C  (high =  +125°C, crit =  +135°C)
PWM Phase4:               +54°C  (high =  +125°C, crit =  +135°C)
PWM Phase5:               +52°C  (high =  +125°C, crit =  +135°C)
CPU FAN:               2760 RPM  (min  300 RPM)
SYS FAN:               1500 RPM  (min  300 RPM)
AUX1 FAN:              1380 RPM  (min  300 RPM)
AUX2 FAN:                 0 RPM  (min  300 RPM)
AUX3 FAN:                 0 RPM  (min  300 RPM)

Re: [gentoo-user] Monitoring temperatures

[Peter Humphrey]

On Friday 20 March 2009 21:14:26 Grant wrote:

> I have another dead power supply and/or another dead motherboard in my
> Gentoo router.  I've tried to make that system as silent as possible
> and I wonder if I'm paying the price.  How do you guys monitor system
> temperatures?  Is lm_sensors the way to go?  How do you keep an eye on
> the temperatures of multiple local and remote systems?

I use gkrellm. Very nice.

-- 
Rgds
Peter

Re: [gentoo-user] Monitoring temperatures

[Dale]

Peter Humphrey wrote:
> On Friday 20 March 2009 21:14:26 Grant wrote:
>
>   
>> I have another dead power supply and/or another dead motherboard in my
>> Gentoo router.  I've tried to make that system as silent as possible
>> and I wonder if I'm paying the price.  How do you guys monitor system
>> temperatures?  Is lm_sensors the way to go?  How do you keep an eye on
>> the temperatures of multiple local and remote systems?
>>     
>
> I use gkrellm. Very nice.
>
>   


Doesn't gkrellm do remote monitoring too?  If so, it has a alert feature
or used to anyway.  You could use that if it still exists to alert you
to high temps.

My personal favorite tho, smart fans.  My CPU has a sensor under it and
varies the CPU fan with temp.  I have done the same with my case fans
and it works pretty well.  One spinning at a good rate is best tho just
to keep air flow at all times.

Dale

:-)  :-) 

Re: [gentoo-user] Monitoring temperatures

[Mike Kazantsev]

[-- Attachment #1: Type: text/plain, Size: 498 bytes --]

On Fri, 20 Mar 2009 18:39:09 -0500
Dale <rdalek1967@gmail.com> wrote:

> My personal favorite tho, smart fans.  My CPU has a sensor under it and
> varies the CPU fan with temp.  I have done the same with my case fans
> and it works pretty well.  One spinning at a good rate is best tho just
> to keep air flow at all times.

No fans will save you when air conditioning system in the data center
goes down! It'd be game over in a couple of minutes ;)

-- 
Mike Kazantsev // fraggod.net

[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 198 bytes --]

Re: [gentoo-user] Monitoring temperatures

[Dale]

Mike Kazantsev wrote:
> On Fri, 20 Mar 2009 18:39:09 -0500
> Dale <rdalek1967@gmail.com> wrote:
>
>   
>> My personal favorite tho, smart fans.  My CPU has a sensor under it and
>> varies the CPU fan with temp.  I have done the same with my case fans
>> and it works pretty well.  One spinning at a good rate is best tho just
>> to keep air flow at all times.
>>     
>
> No fans will save you when air conditioning system in the data center
> goes down! It'd be game over in a couple of minutes ;)
>
>   


Yea, I guess so.  I have forgot to leave my A/C on and it get pretty
warm in here.  All the fans are blasting but it was still running.  Of
course, my little single CPU system running folding is probably not that
big of a deal.

What may also be a good idea is to have some way to shut the systems
down when they get to hot or if the A/C fails.  I would think a
temperature sensor would be better myself.  I !think! gkrellm can do
that.  I'm sure there is a way to do it automatically tho.  After all,
you can't monitor temps 24/7 but the system can.

Dale

:-)  :-) 

Re: [gentoo-user] Monitoring temperatures

[Robin Atwood]

On Saturday 21 Mar 2009, Dale wrote:

> What may also be a good idea is to have some way to shut the systems
> down when they get to hot or if the A/C fails.  I would think a
> temperature sensor would be better myself.  I !think! gkrellm can do
> that.  I'm sure there is a way to do it automatically tho.  After all,
> you can't monitor temps 24/7 but the system can.

If you want to do it professionally nagios is the way to go. 
http://www.nagios.org/
There is an lm_sensors plugin or you can code your own.

HTH
-Robin
-- 
----------------------------------------------------------------------
Robin Atwood.

"Ship me somewheres east of Suez, where the best is like the worst,
 Where there ain't no Ten Commandments an' a man can raise a thirst"
         from "Mandalay" by Rudyard Kipling
----------------------------------------------------------------------

Re: [gentoo-user] Monitoring temperatures

[Neil Bothwick]

[-- Attachment #1: Type: text/plain, Size: 448 bytes --]

On Sat, 21 Mar 2009 02:26:36 -0500, Dale wrote:

> What may also be a good idea is to have some way to shut the systems
> down when they get to hot or if the A/C fails.

Most BIOSes will do this automatically,although they tend to let the
temperature get quite high. lm_sensors includes a daemon that will monitor
temperatures and send warnings and alarms.


-- 
Neil Bothwick

Sure, we just route the main sensor through Data's cat.

[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 198 bytes --]

Re: [gentoo-user] Monitoring temperatures

[Dale]

Neil Bothwick wrote:
> On Sat, 21 Mar 2009 02:26:36 -0500, Dale wrote:
>
>   
>> What may also be a good idea is to have some way to shut the systems
>> down when they get to hot or if the A/C fails.
>>     
>
> Most BIOSes will do this automatically,although they tend to let the
> temperature get quite high. lm_sensors includes a daemon that will monitor
> temperatures and send warnings and alarms.
>
>
>   


I may be wrong here, but doesn't it just shut off like cutting off
power?  Or does it tell the OS to do the shutdown, like in a real
hurry?  I never tested that "feature" so I'm not real clear on how that
works. 

Dale

:-)  :-) 

Re: [gentoo-user] Monitoring temperatures

[Mike Kazantsev]

[-- Attachment #1: Type: text/plain, Size: 1345 bytes --]

On Sat, 21 Mar 2009 17:58:05 -0500
Dale <rdalek1967@gmail.com> wrote:

> Neil Bothwick wrote:
> > On Sat, 21 Mar 2009 02:26:36 -0500, Dale wrote:
> >
> >> What may also be a good idea is to have some way to shut the systems
> >> down when they get to hot or if the A/C fails.
> >
> > Most BIOSes will do this automatically,although they tend to let the
> > temperature get quite high. lm_sensors includes a daemon that will monitor
> > temperatures and send warnings and alarms.
> 
> I may be wrong here, but doesn't it just shut off like cutting off
> power?  Or does it tell the OS to do the shutdown, like in a real
> hurry?  I never tested that "feature" so I'm not real clear on how that
> works. 

Sometimes there is also an option to send ACPI "power button" event
beforehand, but either way that's usually is the "last resort" case -
last thing I want is a hardware shutdown just because of high load.

Besides, it's none too flexible - sometimes just one of the
conditioners goes down, so the room temp gets to, say, 25C, but that's
still not a reason to panic if the situation is under control.

And even when bunch of bioses shut system down all the systems
correctly because of cpu/chipset heat when A/C dies, there'd be a lot of
hard drive failures in a few weeks.

-- 
Mike Kazantsev // fraggod.net

[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 198 bytes --]

Re: [gentoo-user] Monitoring temperatures

[Neil Bothwick]

[-- Attachment #1: Type: text/plain, Size: 951 bytes --]

On Sat, 21 Mar 2009 17:58:05 -0500, Dale wrote:

> > Most BIOSes will do this automatically,although they tend to let the
> > temperature get quite high. lm_sensors includes a daemon that will
> > monitor temperatures and send warnings and alarms.

> I may be wrong here, but doesn't it just shut off like cutting off
> power?  Or does it tell the OS to do the shutdown, like in a real
> hurry?  I never tested that "feature" so I'm not real clear on how that
> works. 

When the temperature gets high enough to trigger a BIOS shutdown, you
want it to happen as fast as possible. If your other safeguards failed to
kick in, either your system is too heavily loaded to shutdown or the
temperature has risen extremely quickly. Either way, cutting the power is
the remaining option.


-- 
Neil Bothwick

Programming Language: (n.) a shorthand way of describing a series of bugs
                      to a computer or a programmer.

[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 198 bytes --]

Re: [gentoo-user] Monitoring temperatures

[Grant]

>> I have another dead power supply and/or another dead motherboard in my
>> Gentoo router.  I've tried to make that system as silent as possible
>> and I wonder if I'm paying the price.  How do you guys monitor system
>> temperatures?  Is lm_sensors the way to go?  How do you keep an eye on
>> the temperatures of multiple local and remote systems?
>
> I use gkrellm. Very nice.
>
> --
> Rgds
> Peter

Thanks everyone, I'll use gkrellm with lm_sensors.

- Grant