[Users] Automatic termination trigger mid-run

[Jordan Nicoules]

Hi,

Thanks, this Cactus parameter was exactly the piece I was missing!

Best,

Jordan
________________________________
From: Users <users-bounces at einsteintoolkit.org> on behalf of Roland Haas via Users <users at einsteintoolkit.org>
Sent: Thursday, June 4, 2026 4:27:40 PM
To: Erik Schnetter via Users
Subject: Re: [Users] Automatic termination trigger mid-run


CUIDADO: Email de um sistema externo. Cuidado com links, anexos e pedidos de dados/senhas.
CAUTION: Email from an external system. Be careful with links, attachments, and requests for data/passwords.

Hello all,

Trigger can set parameters and there's a

Cactus::terminate_next

parameter (src/param.ccl) that you can set.

You may need to then add `Cactus::terminate_next = no`) into your
parfile (so that when the checkpoint recovers the steered value of
terminate_next is overwritten).

Yours,
Roland

> [CAUTION: Non-UBC Email]
>
> Ooh, you are using the "Trigger" thorn. I thought you were using "TerminationTrigger". So that's an option -- use that thorn instead.
>
> Otherwise, if you find your requested feature convenient, then please go ahead and implement it! It would be a nice addition to Cactus.
>
> -erik
>
> > On Jun 4, 2026, at 10:19, Jordan Nicoules <jnicoules at ua.pt> wrote:
> >
> > Hi Erik,
> >
> > Thank you for your reply! I know of CCTK_TerminateNext, I actually use it in one of my thorns (so I could hack it to achieve what I want). In fact, a call to this function is what I would add to Trigger.
> >
> > My point is more: with the existing thorns, is there a way to terminate a simulation while it's running (and before its expected end based on final time/iteration), with a parfile directive, without:
> > - manually editing the termination file
> > - reaching walltime
> > - having NaNs
> > - and such edge cases.
> >
> > Something that would be like:
> > ```
> > Trigger::Trigger_Checked_Variable    [0] = "ADMBase::lapse"
> > Trigger::Trigger_Reduction           [0] = "minimum"
> > Trigger::Trigger_Relation            [0] = "<"
> > Trigger::Trigger_Checked_Value       [0] = 0.2
> > Trigger::Trigger_Reaction            [0] = "terminate"    # <-- This doesn't exist as far as I understand
> > ```
> >
> > Thanks,
> >
> > Jordan
> >
> > From: Users <users-bounces at einsteintoolkit.org <mailto:users-bounces at einsteintoolkit.org>> on behalf of Erik Schnetter via Users <users at einsteintoolkit.org <mailto:users at einsteintoolkit.org>>
> > Sent: Thursday, June 4, 2026 14:57
> > To: users at einsteintoolkit.org <mailto:users at einsteintoolkit.org>
> > Subject: Re: [Users] Automatic termination trigger mid-run
> >
> >
> > CUIDADO: Email de um sistema externo. Cuidado com links, anexos e pedidos de dados/senhas.
> > CAUTION: Email from an external system. Be careful with links, attachments, and requests for data/passwords.
> > You can call the function `CCTK_TerminateNext`, which will stop the simulation gracefully after the current iteration finishes. This is built-in, you do not need a special thorn for this.
> >
> > This is how TerminationTrigger is implemented: It calls this function when a certain file is created.
> >
> > -erik
> >
> >> On Jun 4, 2026, at 09:31, Jordan Nicoules via Users <users at einsteintoolkit.org> wrote:
> >>
> >> Dear all,
> >>
> >> I was wondering if there was a built-in way to trigger the termination of a simulation when something specific happens, along the lines of what Trigger does to steer quantities. From what I understand, Trigger does not offer that possibility, and TerminationTrigger checks either a file or a time limit. What I'm considering would be more along the lines of "an apparent horizon is found so trigger termination", in order to steer some parameters at recovery.
> >>
> >> If there's no such explicit way, I'm happy to try to implement that in Trigger. I actually have ideas of other small improvements to Trigger that I could try contributing as well (namely, increase the maximum number of triggers and allow one trigger to have multiple consequences at once).
> >>
> >> Unfortunately I won't be able to join the weekly call today to discuss this directly.
> >>
> >> Best,
> >>
> >> Jordan
> >> _______________________________________________
> >> Users mailing list
> >> Users at einsteintoolkit.org <mailto:Users at einsteintoolkit.org>
> >> http://lists.einsteintoolkit.org/mailman/listinfo/users
>

--
My email is as private as my paper mail. I therefore support encrypting
and signing email messages. Get my PGP key from http://pgp.mit.edu .
_______________________________________________
Users mailing list
Users at einsteintoolkit.org
http://lists.einsteintoolkit.org/mailman/listinfo/users
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.einsteintoolkit.org/pipermail/users/attachments/20260604/b8fb8595/attachment.htm>