Next post, how I recovered the database of SSSD when I deleted the whole path /var/lib/sss/ by accident. ๐ฅฒ
22.10.2025 20:59 โ ๐ 0 ๐ 0 ๐ฌ 0 ๐ 0@ogmdevlab.bsky.social
Computational Geophysicist | Finite Element Analysis | Geothermal Reservoir Modeling | HPC | WEB & Back End Tools, Ph.D. Student at @ub.edu @marcusfinix@sciences.social ๐ฆฃ
Next post, how I recovered the database of SSSD when I deleted the whole path /var/lib/sss/ by accident. ๐ฅฒ
22.10.2025 20:59 โ ๐ 0 ๐ 0 ๐ฌ 0 ๐ 0Moreover, to enable secure external access, I configured a bridge acting as a switch to assign static IPs to the login node and a bastion node.
The bastion node will verify and admit external connections, enforcing two-factor authentication for enhanced security. All of this powered by #freeIPA
Additionally, I have decided to isolate the computing nodes network from the university network as a very basic attempt to reduce latency in message communications.
Something more or less like this ๐๐ผ
Cluster Update
We are nearing the completion of the restart process for all services. Here is the current status of the login node services:
๐ข FreeIPA
โช๏ธ LDAP
โช๏ธ SSSD
โช๏ธ Kerberos
๐ข Munge
๐ก SLURM
๐ข AutoNFS
๐ข Modules
๐ด Grafana
๐ด Prometheus
By the way, the fortran post ๐๐ผ
bsky.app/profile/ogmd...
I remember I prepared this nice ilustration about #fortran but I neves share it.๐
Well... here it go
However, I also need to remember that last July month, marked one year since I was forced to leave my previous PhD program ๐. That unfortunate event deserves its own post, and maybe Iโll start sharing that story soon...๐ฅบ
#phdlife
The last time I posted here, I shared some thoughts about #Fortran, and I intended to talk about my #PhD journey. The first year has been tough, but I managed to overcome many challenges, from field work to #CAD modeling, I complete tasks satisfactorily.
22.10.2025 00:04 โ ๐ 1 ๐ 0 ๐ฌ 1 ๐ 0Otro dia a la espera de que llegue el SSD que necesito para restablecer el nodo login.
Pienso hacer un post cortito sobre el proceso y otro mas detallado, allรก en #mastodon ๐ฆฃ
Despuรฉs de dejar listo el clรบster MAGMA, montado a base de mucho ensayo, error y afinado de Slurm, hoy el nodo login dijo basta. Todavรญa corrรญa sobre un HDD que ya habrรก cumplido su ciclo. Ahora toca retroceder un paso, montar SSD y RAM nuevos, y restaurar el acceso al clรบster. ๐ง๐ป
28.09.2025 15:26 โ ๐ 0 ๐ 0 ๐ฌ 0 ๐ 0Hoy retome mis clases de Catalรกn, despuรฉs de que la pandemia me arruinara mi primer curso oficial, hace ya 5 aรฑos. Espero terminarlo esta vez, sin ningรบn cisne negro
16.09.2025 18:32 โ ๐ 0 ๐ 0 ๐ฌ 0 ๐ 0Apenas llegue a esta red y me ausente un tiempo. De un momento a otro me desborde en deberes y actividades propias del doctorado. Pero ahora estoy aqui con mucho que contar y compartir.
13.09.2025 14:04 โ ๐ 0 ๐ 0 ๐ฌ 0 ๐ 0Magmatic intrusions look amazing ๐
#canaryislands
Another day at #CanaryIslands field work, preparing the #magnetoteluric sounding wire conectionsโฆ
I know, it seems like I know what I am doing ๐
#geophysics #vulcanology
This week I'm out doing field work, and it's taking me more time than I expected to share a new thread about computational #electromagnetics here soon.
I hope to post it shortly
#CEM
I decided to split my post btw ๐ฆ and ๐ฆฃ might lean toward more technical content on Mastodon:
[ sciences.social/@marcusfinix ]
...or maybe not ๐
---
Are you on Mastodon too?
๐๐ค
26.04.2025 21:15 โ ๐ 2 ๐ 0 ๐ฌ 0 ๐ 0Hey Bruno!
Yeah he brings great contributions to Fortran language and community.
On the other hand, I hope there in BSC, you will have the opportunity to use codes like Alya to work with Fortran.
For the past 3 months, Iโve been attending an #NPTEL course on Scientific #HPC, with lectures delivered by IIT Bombay.
Tomorrow is the big day: my proctored exam!
Iโve studied as much as I could. ๐ฐ
Letโs see if I can earn the certification.
One week left to apply for the EUMaster4HPC Summer School 2025! From 14โ25 July at METU, Ankara (TR), explore #HPC for specialised applications in Data Science, Chemistry, Aeronautics & more. Includes expert lectures, industry insights, and social events. Learn more: eumaster4hpc.eu/summer-schoo...
23.04.2025 10:27 โ ๐ 4 ๐ 1 ๐ฌ 0 ๐ 0As a #geophysicist.
Iโve done fieldwork in ERT, GPR, MASW, and VES but always avoided magnetotellurics. It just didnโt click with me.
Now life is funny: in a few days, Iโll be doing my first MT surveys, guided by my #PhD advisor.
Sometimes, the path you resist is exactly where youโre meant to grow
Thank you!
Yeah I responded the email. Also I received and checked the next ones sent by you, thank you so much!
I just discovered your account, and I'm very interested in attending the course. The registration is now closed, but I sent an email asking for a slot. Is there any chance I could get one?
24.04.2025 15:57 โ ๐ 2 ๐ 0 ๐ฌ 1 ๐ 0In the end, modern languages like ๐ have taken most of the attention in the scientific community.
However, Fortran is still alive and continuously evolving. With great community.
Whatโs your take on the โold vs. modernโ. Have you worked with Fortran too?
[5/5]
Thanks for reading :)
Since then, I've stood by this: Fortran is not dead. It updates, improves, and stays relevant especially in scientific computing where raw performance matters.
It may not be the prettiest language (in fact it is) , but itโs robust, fast, and efficient. It gets the job done without drama.
[4/5]
I first encountered Fortran during my undergrad studies (2010). Many years later, I picked it up again to develop a 2.5D transient electromagnetic modeling code for geophysical applications. I worked on it for the last 4 years and that project shaped the way I approach simulation today.
[3/5]
* Object-oriented code
* Use modules and submodules
* Implement native parallelism with coarrays
* Build #HPC algorithms.
* Write #CUDA codes with #CudaFortran
* Target modern architectures with performance in mind.
* Even some #ANN implementation are based on this amazing language
[2/5]
Let's talk about #Fortran and why so many people think of it as a old language. Maybe because it started in the 1950s. Or maybe because it doesnโt look like Python.
But hereโs the thing: Fortran has evolved.
Today, you can write code such as:
[1/5]
Me han dado unas ganas irresistibles por volver a construir en #C++
Comenzar con, por ejemplo,
Exportar mi codigo #FEM en 2.5D de #FORTRAN a #C++
๐คฉ
Recien aterrizando ๐ฆ
PROGRAM BLUESKY
PRINT *, "Exploring a new decentralized future..."
END
Hello bluesky!
Veamos de que va esto ๐ญ