Posted in 2026
LiteLLM python package compromised
- 27/03/2026
On March 24, 2026, the popular LiteLLM python package was compromised by a supply chain attack. If you are using LiteLLM in any form, please take immediate action and check if any of the compromised versions (v1.82.7 or v1.82.8) are installed. See the LiteLLM blog post on the incident for all the details on how to check if you are affected and what to do if that’s the case: https://docs.litellm.ai/blog/security-update-march-2026.
Improved way to launch interactive jobs
- 24/03/2026
We’ve changed the recommended way to launch an interactive job on a compute
node from the terminal interface. Previously, this was done using srun --pty
bash -l. We now recommend using salloc, which makes it possible to launch
parallel MPI jobs with srun or mpirun from within the interactive
session. All related documentation has been updated accordingly, see the
section on Interactive jobs for details.
VSC Web Authentication Outage
- 18/03/2026
10:00 The authentication to VSC web services has been restored. All users from VUB and UZB can access our services normally.
Update of the HPC portal
- 27/02/2026
We have updated the VUB OnDemand portal by installing several new apps, upgrading to Open OnDemand 4.1 and making a slight change to the job submission forms.
VUB HPC Overview 2025
- 23/02/2026
2025 marked a year of strong growth and strategic preparation for the SDC team. While significant effort was invested in preparing for the next VSC Tier-1 Sofia , the Tier-2 infrastructure (Hydra and Anansi) continued to expand in usage, users, and services.
VUB-HPC briefly offline on 23 February
- 18/02/2026
Due to a scheduled firewall migration, the VUB-HPC clusters will be temporarily inaccessible on 23 February. This should not take longer than a few minutes. Users will be unable to access the clusters during the migration. Running jobs will continue to run, provided they do not require external internet access.
Network upgrade on scratch
- 17/02/2026
14:00 The upgrade operation has been completed ahead of schedule. All new components were successfully installed. The job queue in Hydra is again accepting jobs.
18,20/3/2026: Linux and HPC Introduction Trainings
- 10/02/2026
We are pleased to announce on-site Linux and HPC introductions on the VUB Health campus in Jette, in cooperation with the Vlaams Supercomputer Centrum (VSC).
New GPU nodes added to Hydra
- 05/02/2026
We’re happy to announce the public availability of hopper_gpu, a brand new GPU partition in Hydra. This partition features 5 nodes, each equipped with 2 NVIDIA H200 GPUs (140 GB VRAM) and 48 CPU cores.
User limits on Anansi
- 04/02/2026
We have added new limits to the number of CPU cores and GPU resources a single user can request across all jobs on the interactive/debug cluster Anansi. The aim is to improve queueing times and to avoid situations where a single user submits many jobs and uses up the resources of Anansi, effectively blocking other users from starting their session on OnDemand or batch jobs.
Upgrade to Rocky Linux 9.7
- 27/01/2026
We are upgrading the OS and Slurm scheduler on Hydra and Anansi starting Tuesday, January 27. The rollout will start Tuesday night. User impact is minimal: there will be no downtime, and your running or queued jobs will remain unaffected.