Mailing List Archives

Public Access

 		[Computer Systems Lab]

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [HTCondor-users] Power management not working in Windows


No, I don't have any configuration for STARTD_ATTRS or STARTD_EXPRS:

> condor_config_val -v -dump STARTD_ATTRS
# Configuration from machine: ZARATHUSTRA

# Parameters with names that match STARTD_ATTRS:
STARTD_ATTRS =
Â# at: <Default>
Â# expanded:
SYSTEM_STARTD_ATTRS = COLLECTOR_HOST_STRING DedicatedScheduler
Â# at: <Default>
Â# expanded: COLLECTOR_HOST_STRING DedicatedScheduler
Â# default: COLLECTOR_HOST_STRING DedicatedScheduler

> condor_config_val -v -dump STARTD_EXPRS
# Configuration from machine: ZARATHUSTRA

# Parameters with names that match STARTD_EXPRS:

> condor_config_val -dump | grep -I hiber
ROOSTER_UNHIBERNATE = Offline && Unhibernate

> condor_config_val -dump | grep -i hiber
HIBERNATE = ifThenElse($(ShouldHibernate), $(HibernateState), 0)
HIBERNATE_CHECK_INTERVAL = 300> condor_status -long | grep -i hiber
CanHibernate = true
HibernationLevel = 0
HibernationState = "NONE"
HibernationSupportedStates = "S3,S4,S5"
Unhibernate = MY.MachineLastMatchTime =!= undefined
HibernateState = 3
ROOSTER_MAX_UNHIBERNATE = 0
ROOSTER_UNHIBERNATE = Offline && Unhibernate
ROOSTER_UNHIBERNATE_RANK =
ShouldHibernate = ( (KeyboardIdle > $(StartIdleTime)) && $(CPUIdle) && ($(StateTimer) > $(TimeToWait)) )

> condor_status -long | findstr -i hibernate
CanHibernate = true
Unhibernate = MY.MachineLastMatchTime =!= undefined

But:

> condor_status -long | grep -i hiber
CanHibernate = true
HibernationLevel = 0
HibernationState = "NONE"
HibernationSupportedStates = "S3,S4,S5"
Unhibernate = MY.MachineLastMatchTime =!= undefined

So it looks like the configuration is OK...I am puzzled...

Best

On Fri, May 18, 2018 at 1:40 PM, John M Knoeller <johnkn@xxxxxxxxxxx> wrote:

maybe something is overwriting that attribute?ÂÂ do you have configuration for ÂSTARTD_ATTRS or STARTD_EXPRS?

Â

condor_config_val -v -dump STARTD_

Â

or maybe

Â

condor_config_val -dump | grep -I hiber

Â

This is what I see when I run condor_status

Â

> condor_status -long | grep -I hiber

CanHibernate = true

HibernationLevel = 0

HibernationState = "NONE"

HibernationSupportedStates = "S3,S4,S5"

Unhibernate = MY.MachineLastMatchTime =!= undefined

.. repeats multiple timesâ

Â

Â

Â

-tj

Â

From: HTCondor-users [mailto:htcondor-users-bounces@xxxxxxxxxxx] On Behalf Of Julio Augusto Freyre-Gonzalez
Sent: Friday, May 18, 2018 12:57 PM


To: HTCondor-Users Mail List <htcondor-users@xxxxxxxxxxx>
Subject: Re: [HTCondor-users] Power management not working in Windows

Â

Ok. Yes, in fact, most of the time the queues are empty. I configured

Â

STARTD_DEBUG = $(STARTD_DEBUG) D_CAT D_ALWAYS:2Â

Â

and found the following:

Â

05/18/18 12:39:29 HibernationManager: Hibernation is enabled
05/18/18 12:47:41 (D_ALWAYS:2) Canceled hibernation timer (31)
05/18/18 12:47:53 (D_ALWAYS) HibernationManager: Hibernation is enabled
05/18/18 12:47:53 (D_ALWAYS:2) Using network interface for hibernation
05/18/18 12:47:53 (D_ALWAYS:2) Initially invoking hibernation plugin 'D:\Condor\bin/condor_power_state ad'
05/18/18 12:47:53 (D_ALWAYS:2) Detected hibernation states: S3,S4,S5

Â

So, Startd is finding the hibernation states but a

Â

condor_status -long | findstr -I hibernate

Â

only shows

Â

CanHibernate = true
Unhibernate = MY.MachineLastMatchTime =!= undefined

Â

so there is no HibernationSupportedStates in the slot ads.

Â

any idea?

Â

Â

On Fri, May 18, 2018 at 12:17 PM, John M Knoeller <johnkn@xxxxxxxxxxx> wrote:

Actually, if all of your machines are also SCHEDDâs, then they probably always have empty queues. A SCHEDD with an empty queue is OK. but ONLY a Startd will pay attention to the HIBERNATE _expression_.

So a SCHEDD only machine will never *choose* to hibernate.

Â

I think your real problem may be that the Startd doesnât believe that S3 hibernation is available, if it did you would see that in the condor_status -long output.

Â

From: HTCondor-users [mailto:htcondor-users-bounces@xxxxxxxxxxx] On Behalf Of Julio Augusto Freyre-Gonzalez
Sent: Friday, May 18, 2018 10:33 AM


To: HTCondor-Users Mail List <htcondor-users@xxxxxxxxxxx>
Subject: Re: [HTCondor-users] Power management not working in Windows

Â

Aha! Thanks John. The problem I think has to do with the fact that all the machines are able to submit (all have a SCHEDD). I will remove the SCHEDD in one of them to carry out some test.

Â

Best

--
Dr. Julio Augusto Freyre-Gonzalez (PhD)
Associate Professor
Head of the Regulatory Systems Biology Group
http://freyrelab.org
Evolutionary Genomics Program
Center for Genomic Sciences
National Autonomous University of Mexico (UNAM)
Av. Universidad s/n, Col. Chamilpa, 62210
Cuernavaca, Morelos. Mexico

Â

El vie., 18 de may. de 2018 10:02, John M Knoeller <johnkn@xxxxxxxxxxx> escribiÃ:

The HTCondor Central manager will never attempt to hibernate. ÂÂOnly a machine that has an HTCondor STARTD, but not a SCHEDD, NEGOTIATOR, or COLLECTOR is a candidate for hibernation.

Â

Also, hibernation can be rejected by processes outside of HTCondor.ÂÂ

Â

Â

If you configure

Â

STARTD_DEBUG = $(STARTD_DEBUG) D_CAT D_ALWAYS:2

Â

To see more output in the StartdLog.

Â

You should see something like

Â

05/16/18 13:53:10 (D_ALWAYS:2) Detected hibernation states: S3

Â

in the StartdLog

Â

In addition to

ÂÂÂCanHibernate = true
ÂÂ Unhibernate = MY.MachineLastMatchTime =!= undefined

Â

You should also be seeing an attribute HibernationSupportedStates in the slot ads.

Â

Â

-tj

Â

Â

Â

Â

Â

Â

Â

HIBERNATE = âRAMâ on ALL slots in the Startd, then youd

Â

From: HTCondor-users [mailto:htcondor-users-bounces@xxxxxxxxxxx] On Behalf Of Julio Augusto Freyre-Gonzalez
Sent: Thursday, May 17, 2018 6:31 PM
To: HTCondor-Users Mail List <htcondor-users@xxxxxxxxxxx>
Subject: Re: [HTCondor-users] Power management not working in Windows

Â

I even used HIBERNATE = "RAM" to force sleep (using HIBERNATE_CHECK_INTERVAL = 300), but computer does not sleep.

Â

condor_config_val -v Hibernate

Â

HIBERNATE = "RAM"
Â# at: D:\Condor\etc\condor_config.power, line 40
Â# raw: HIBERNATE = "RAM"

Â

condor_config_val -v Hibernate_check_interval

Â

HIBERNATE_CHECK_INTERVAL = 300
Â# at: D:\Condor\etc\condor_config.power, line 17
Â# raw: HIBERNATE_CHECK_INTERVAL = 300

Â

Best

Â

On Thu, May 17, 2018 at 6:24 PM, Julio Augusto Freyre-Gonzalez <jfreyre@xxxxxxxxxxx> wrote:

John,

condor_status -long | findstr -I hibernate shows:

Â

CanHibernate = true
Unhibernate = MY.MachineLastMatchTime =!= undefined
CanHibernate = true
Unhibernate = MY.MachineLastMatchTime =!= undefined
CanHibernate = true
Unhibernate = MY.MachineLastMatchTime =!= undefined

.

. (the same for all the 74 slots)

.

CanHibernate = true
Unhibernate = MY.MachineLastMatchTime =!= undefined

Â

Additionally, cat *.* | grep -i powereventhander ran on log directory shows:

Â

05/16/18 13:53:10 PowerEventHander: Machine entering hibernation
05/16/18 15:14:30 PowerEventHander: Waking machine to handle an event (Automatic)
05/16/18 15:14:30 PowerEventHander: Waking machine (APM)
05/17/18 02:56:39 PowerEventHander: Machine entering hibernation
05/17/18 11:41:14 PowerEventHander: Waking machine to handle an event (Automatic)
05/17/18 11:41:18 PowerEventHander: Waking machine (APM)
05/17/18 15:59:24 PowerEventHander: Machine entering hibernation
05/17/18 18:14:21 PowerEventHander: Waking machine to handle an event (Automatic)
05/17/18 18:14:25 PowerEventHander: Waking machine (APM)

Â

and > cat *.* | grep -i hibernation


05/16/18 13:53:10 PowerEventHander: Machine entering hibernation
05/17/18 02:56:39 PowerEventHander: Machine entering hibernation
05/17/18 15:59:24 PowerEventHander: Machine entering hibernation
05/16/18 15:14:37 HibernationManager: Hibernation is enabled
05/16/18 20:42:56 HibernationManager: Hibernation is enabled
05/16/18 21:57:24 HibernationManager: Hibernation is enabled
05/17/18 11:41:22 HibernationManager: Hibernation is enabled
05/17/18 14:51:46 HibernationManager: Hibernation is enabled

Â

But the 'Waking machine' events were trigger by myself waking up the computer, I configured this machine to sleep after a timeout using the control panel. But the other machines, including the central manager, are never going to sleep.

Â

Best

Â

On Thu, May 17, 2018 at 5:17 PM, John M Knoeller <johnkn@xxxxxxxxxxx> wrote:

what does

Â

condor_status -long | findstr -I hibernate

Â

show?

Â

Â

From: HTCondor-users [mailto:htcondor-users-bounces@xxxxxxxxxxx] On Behalf Of Julio Augusto Freyre-Gonzalez
Sent: Thursday, May 17, 2018 12:00 PM
To: HTCondor-Users Mail List <htcondor-users@xxxxxxxxxxx>
Subject: [HTCondor-users] Power management not working in Windows

Â

I am trying to enable power magement in condor for Windows. I am using the condor_config.power available in condor/etc/. I put a 'include : $(RELEASE_DIR)\etc\condor_config.power' in condor_config. Also, I enabled ROOSTERÂ and put a 'COLLECTOR_PERSISTENT_AD_LOG = $(RELEASE_DIR)\log\PersistentAds' in the central manager. But after 36hrs computers keep awake. I already verified that S3 mode is supported in al computers using 'powercfg -A'.

Â

--

Dr. Julio Augusto Freyre-Gonzalez (PhD)
Associate Professor
Head of the Regulatory Systems Biology Group

Evolutionary Genomics Program
Center for Genomic Sciences
National Autonomous University of Mexico (UNAM)

Â

Â

_______________________________________________
HTCondor-users mailing list
To unsubscribe, send a message to htcondor-users-request@cs.wisc.edu with a
subject: Unsubscribe
You can also unsubscribe by visiting
https://lists.cs.wisc.edu/mailman/listinfo/htcondor-users

The archives can be found at:
https://lists.cs.wisc.edu/archive/htcondor-users/




--

Dr. Julio Augusto Freyre-Gonzalez (PhD)
Associate Professor
Head of the Regulatory Systems Biology Group
Evolutionary Genomics Program
Center for Genomic Sciences
National Autonomous University of Mexico (UNAM)
Av. Universidad s/n, Col. Chamilpa, 62210
Cuernavaca, Morelos. Mexico
http://orcid.org/0000-0001-7061-7637
http://unam.academia.edu/jfreyre
https://www.researchgate.net/profile/Julio_Augusto_Freyre-Gonzalez/

"Ser joven y no ser revolucionario es una contradiccion hasta biologica"
- Salvador Allende Gossens (1908-1973)

"If we knew what it was we were doing, it would not be called research, would it?"
- Albert Einstein (1879-1955)

"There is one thing even more vital to science than intelligent methods; and that is, the sincere desire to find out the truth, whatever it may be"
- Charles Sanders Pierce (1839-1914)

"The great tragedy of science -- the slaying of a beautiful hypothesis by an ugly fact"
- Thomas Henry Huxley (1825-1895)




--

Dr. Julio Augusto Freyre-Gonzalez (PhD)
Associate Professor
Head of the Regulatory Systems Biology Group
Evolutionary Genomics Program
Center for Genomic Sciences
National Autonomous University of Mexico (UNAM)
Av. Universidad s/n, Col. Chamilpa, 62210
Cuernavaca, Morelos. Mexico
http://orcid.org/0000-0001-7061-7637
http://unam.academia.edu/jfreyre
https://www.researchgate.net/profile/Julio_Augusto_Freyre-Gonzalez/

"Ser joven y no ser revolucionario es una contradiccion hasta biologica"
- Salvador Allende Gossens (1908-1973)

"If we knew what it was we were doing, it would not be called research, would it?"
- Albert Einstein (1879-1955)

"There is one thing even more vital to science than intelligent methods; and that is, the sincere desire to find out the truth, whatever it may be"
- Charles Sanders Pierce (1839-1914)

"The great tragedy of science -- the slaying of a beautiful hypothesis by an ugly fact"
- Thomas Henry Huxley (1825-1895)

_______________________________________________
HTCondor-users mailing list
To unsubscribe, send a message to htcondor-users-request@cs.wisc.edu with a
subject: Unsubscribe
You can also unsubscribe by visiting
https://lists.cs.wisc.edu/mailman/listinfo/htcondor-users

The archives can be found at:
https://lists.cs.wisc.edu/archive/htcondor-users/


_______________________________________________
HTCondor-users mailing list
To unsubscribe, send a message to htcondor-users-request@cs.wisc.edu with a
subject: Unsubscribe
You can also unsubscribe by visiting
https://lists.cs.wisc.edu/mailman/listinfo/htcondor-users

The archives can be found at:
https://lists.cs.wisc.edu/archive/htcondor-users/




--

Dr. Julio Augusto Freyre-Gonzalez (PhD)
Associate Professor
Head of the Regulatory Systems Biology Group
Evolutionary Genomics Program
Center for Genomic Sciences
National Autonomous University of Mexico (UNAM)
Av. Universidad s/n, Col. Chamilpa, 62210
Cuernavaca, Morelos. Mexico
http://orcid.org/0000-0001-7061-7637
http://unam.academia.edu/jfreyre
https://www.researchgate.net/profile/Julio_Augusto_Freyre-Gonzalez/

"Ser joven y no ser revolucionario es una contradiccion hasta biologica"
- Salvador Allende Gossens (1908-1973)

"If we knew what it was we were doing, it would not be called research, would it?"
- Albert Einstein (1879-1955)

"There is one thing even more vital to science than intelligent methods; and that is, the sincere desire to find out the truth, whatever it may be"
- Charles Sanders Pierce (1839-1914)

"The great tragedy of science -- the slaying of a beautiful hypothesis by an ugly fact"
- Thomas Henry Huxley (1825-1895)


_______________________________________________
HTCondor-users mailing list
To unsubscribe, send a message to htcondor-users-request@cs.wisc.edu with a
subject: Unsubscribe
You can also unsubscribe by visiting
https://lists.cs.wisc.edu/mailman/listinfo/htcondor-users

The archives can be found at:
https://lists.cs.wisc.edu/archive/htcondor-users/



--
Dr. Julio Augusto Freyre-Gonzalez (PhD)
Associate Professor
Head of the Regulatory Systems Biology Group
Evolutionary Genomics Program
Center for Genomic Sciences
National Autonomous University of Mexico (UNAM)
Av. Universidad s/n, Col. Chamilpa, 62210
Cuernavaca, Morelos. Mexico
http://orcid.org/0000-0001-7061-7637
http://unam.academia.edu/jfreyre
https://www.researchgate.net/profile/Julio_Augusto_Freyre-Gonzalez/

"Ser joven y no ser revolucionario es una contradiccion hasta biologica"
- Salvador Allende Gossens (1908-1973)

"If we knew what it was we were doing, it would not be called research, would it?"
- Albert Einstein (1879-1955)

"There is one thing even more vital to science than intelligent methods; and that is, the sincere desire to find out the truth, whatever it may be"
- Charles Sanders Pierce (1839-1914)

"The great tragedy of science -- the slaying of a beautiful hypothesis by an ugly fact"
- Thomas Henry Huxley (1825-1895)