Re: [HTCondor-users] 'condor_off -peaceful' kills jobs

[Brian Bockelman <bbockelm@xxxxxxxxxxx>]

Hi,

Basically, âcondor_off -peacefulâ sends two logical commands:

1) âTurn off gracefullyâ.
2) âUpgrade graceful off to peaceful offâ.

IIRC, itâs possible to configure HTCondor security settings so (1) succeeds and (2) fails (of course, there may be other reasons (2) fails to be successfully delivered).  I believe TJ is referring to a future effort to combine these into one atomic command that either fully-succeeds or fully-fails.

Brian

> On Feb 24, 2016, at 12:36 PM, Bob Ball <ball@xxxxxxxxx> wrote:
> 
> Interestingly, just to throw some more rocks onto this pile, we have never had a problem with this command, issued as "root" from our condor master machine:
> condor_off -peaceful -daemon startd -name $publicName
> where publicName is the FQDN under which Condor knows the machine. This is the 8.2 series on Linux, and earlier.  We just went to 8.4, and so far I've no experience with it.
> 
> Perhaps there is some odd security issue intervening as well?
> 
> bob
> 
> 
> On 2/24/2016 1:21 PM, John M Knoeller wrote:
>> The HTCondor team will be the lookout this bug ticket, we had already been planning to make some changes to the way peaceful shutdown is delivered, this is yet another reason why we would want to do that.
>> 
>> In the meantime, I would suggest that instead of condor_off -peaceful   you could use condor_drain to drain the node
>> And then condor_off to shut it off once it has drained.    condor_drain doesn't have a peaceful mode, but it does
>> honor MaxVacateTime and MaxJobRetirementTime while draining.
>> 
>> -tj
>> 
>> -----Original Message-----
>> From: HTCondor-users [mailto:htcondor-users-bounces@xxxxxxxxxxx] On Behalf Of Di Qing
>> Sent: Wednesday, February 24, 2016 11:35 AM
>> To: HTCondor-Users Mail List <htcondor-users@xxxxxxxxxxx>
>> Subject: Re: [HTCondor-users] 'condor_off -peaceful' kills jobs
>> 
>> Ben,
>> 
>> Thanks. I did not create the ticket, please go ahead to put the ticket.
>> 
>> Thanks,
>> 
>> Di
>> 
>> On 24/02/16 06:03 AM, Ben Cotton wrote:
>>> Di,
>>> 
>>> If it makes you feel any better, I recently saw the same behavior on
>>> 8.2.9, but didn't have time to look at it in depth. If you don't, I'll
>>> put a ticket in the HTCondor issue tracker later today. I have some
>>> log snippets that show it's clearly doing the wrong thing.
>>> 
>>> 
>>> Thanks,
>>> BC
>>> 
>>> 
>> _______________________________________________
>> HTCondor-users mailing list
>> To unsubscribe, send a message to htcondor-users-request@xxxxxxxxxxx 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@xxxxxxxxxxx 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@xxxxxxxxxxx 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/