[ase-users] SocketIOCalculator i.c.w. NEB

[Ask Hjorth Larsen]

Dear Willem,

Am Do., 9. Jan. 2020 um 13:51 Uhr schrieb Offermans Willem <
willem.offermans at vito.be>:

> Dear Ask,
>
> Thank you for your quick response. It is appreciated.
>
> By trial and error, I already figured out that each image needs its own
> SocketIOCalculator and moreover its own unixsocket.
>
> What do you mean with different port for each image?
> Does ``port`` need to be used if server and client are on different
> machines?
> Does ``socket`` need to be used if server and client are on the same
> machine?
> Are socket and port mutually exclusive?
>

Ah, ignore port if you use unix sockets.


>
> Parallelisation is an open issue, since I’m running on a computer cluster
> without mpi.
> The maintainers don’t want/cannot install mpirun.
>

You can probably use multiprocessing then, but let's ignore that for now.


>
> For the moment I’m running NEB on a single client in a serial fashion.
>
> Could I parallelise over images without mpi?
>

Yes, with threading.  NEB has that feature but it may be finicky.


> It should be possible in principle, but I don’t know how to do this for
> now.
>
> B.T.W. If I understand the concept correctly, the description of the
> SocketIOCalculator on the wiki is not correct.
>
> See: https://wiki.fysik.dtu.dk/ase/ase/calculators/socketio/socketio.html
>
> ``Some codes can run in “driver mode” where a server provides atomic
> coordinates through a socket connection, and the code returns energies,
> forces, and stress to the server. That way the startup overhead is
> eliminated, and the codes can reuse and extrapolate wavefunctions and other
> quantities for increased efficiency.
>
> ASE provides such a server in the form of a calculator.``
>
> Should it be?:
>
> ``Some codes can run in “driver mode” (server mode) where a client
> provides atomic coordinates through a socket connection to the server, and
> the server returns energies, forces, and stress to the client. That way the
> startup overhead is eliminated, and the server can reuse and extrapolate
> wavefunctions and other quantities for increased efficiency. It represents
> a classical client server setup.
>
>
No, the computational code runs as client.  The SocketIOCalculator is the
server.

User -> SocketIOCalculator -> server -> socket -> computational code

Best regards
Ask


> ASE provides such a server in the form of a calculator, i.e. the
> SocketIOCalculator.``
>
>
>
>
> Met vriendelijke groeten,
> Mit freundlichen Grüßen,
> With kind regards,
>
>
> Willem Offermans
> Researcher Electrocatalysis SCT
> VITO NV | Boeretang 200 | 2400 Mol
> Phone:+32(0)14335263 Mobile:+32(0)492182073
>
> Willem.Offermans at Vito.be
>
>
> On 9 Jan 2020, at 13:16, Ask Hjorth Larsen <asklarsen at gmail.com> wrote:
>
> Dear Willem,
>
> Am Do., 9. Jan. 2020 um 11:50 Uhr schrieb Offermans Willem via ase-users <
> ase-users at listserv.fysik.dtu.dk>:
>
>> Dear ASE friends,
>>
>> Is it useful to use the SocketIOCalculator in combination with a NEB
>> calculation?
>>
>
> Definitely, as long as each image has one SocketIOCalculator.
>
>
>>
>>
>> Is it possible to use the SocketIOCalculator in combination with a NEB
>> calculation?
>>
>
> Probably!  Use a different port for each image.
>
> It could be done more conveniently by adding protocol support multiple
> images on a single server (called 'beads' in i-py), but that won't happen
> just now unless someone does it. :)
>
> Things will also depend on how you want to parallelize.
>
>
>>
>>
>> Is there any example available to describe how to use above combination?
>>
>
> No, but one would be much appreciated.
>
>  * Write a working relaxation with socketio
>  * Write/grab a working NEB with EMT
>  * Try to combine the two
>
> Best regards
> Ask
>
>
>>
>>
>>
>> Met vriendelijke groeten,
>> Mit freundlichen Grüßen,
>> With kind regards,
>>
>>
>> Willem Offermans
>> Researcher Electrocatalysis SCT
>> VITO NV | Boeretang 200 | 2400 Mol
>> Phone:+32(0)14335263 Mobile:+32(0)492182073
>>
>> Willem.Offermans at Vito.be
>>
>> <vito.jpg>
>>
>> Indien u VITO Mol bezoekt, hou aub er dan rekening mee dat de hoofdingang
>> voortaan enkel bereikbaar is vanuit de richting Dessel-Retie, niet vanuit
>> richting Mol, zie vito.be/route. <http://www.vito.be/route>
>> If you plan to visit VITO at Mol, then please note that the main entrance
>> can only be reached coming from Dessel-Retie and no longer coming from Mol,
>> see vito.be/en/contact/locations.
>> <http://www.vito.be/en/contact/locations>
>> VITO Disclaimer: http://www.vito.be/e-maildisclaimer
>> _______________________________________________
>> ase-users mailing list
>> ase-users at listserv.fysik.dtu.dk
>> https://listserv.fysik.dtu.dk/mailman/listinfo/ase-users
>
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://listserv.fysik.dtu.dk/pipermail/ase-users/attachments/20200109/d06d173a/attachment-0001.html>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: vito.jpg
Type: image/jpeg
Size: 15232 bytes
Desc: not available
URL: <http://listserv.fysik.dtu.dk/pipermail/ase-users/attachments/20200109/d06d173a/attachment-0001.jpg>