[Nut-upsuser] ups.delay.shutdown clarification

[Des Dougan]

The nut version is nut-2.2.0-3.3.el4, running on Mitel SME Server 7.3  
(based on CentOS 4.6).

Des

On 25 Feb 2009, at 11:31, Arjen de Korte wrote:

> Citeren Des Dougan <des at DouganConsulting.com>:
>
>> A client of mine has an APC Back-UPS 350 model (USB connected). His
>> office had a power incident yesterday, where there was an  
>> interruption
>> of about 30 seconds. During that time, the server shut down. As soon
>> as it went on battery, the daemon initiated a shutdown. In checking  
>> it
>> today, I see the ups.delay.shutdown setting is -1. I tried to adjust
>> it, but could not do so.
>>
>> What does the "-1" setting mean?
>
> It means no shutdown delay value has been configured. I already  
> wrote that we'll need the NUT version and OS to determine what  
> *exactly* is going on.
>
>> I Googled, but couldn't find anything
>> relevant. Am I correct in assuming it is this setting that caused the
>> immediate shutdown?
>
> No, the decision to shut the system down is unrelated to the  
> ups.delay.shutdown value. Most likely (but see the note above), the  
> UPS firmware is broken and it signals low battery right after the  
> power is lost. NUT has no other option than to initiate system  
> shutdown in that case. Note that the Back-UPS 350 is a *very* small  
> entry level UPS that really shouldn't be used for anything other  
> than small desktop systems.
>
>> Today, the UPS shows a load of 35, via the upsc
>> command, so it doesn't appear it was overloaded.
>
> Maybe, maybe not. The only way to conclusively tell why NUT decided  
> to shutdown the system is to run the driver in debug mode. It will  
> probably confirm the above.
>
> Best regards, Arjen
> -- 
> Please keep list traffic on the list
>

--

Des Dougan, Principal
Dougan Consulting Group Inc.

Office:		604-628-5434
Cell:   	604-866-2848
Email:  	des at DouganConsulting dot com

         www.DouganConsulting.com

Peace of Mind, One Computer at a Time.