News from RCIC and elsewhere

HPC3/HPC2 Downtime Aug 30, 2022s

Monday, Aug 01 2022 by Philip Papadopoulos

A Major OS Update will begin at 8am, on Tuesday, August 30, 2022

This is Phase 2 of the upgrade (Phase 1 was June 15, 2022).

HPC2 and HPC3 will upgraded to Enterprise Linux 8 (Rocky Linux) from the current CentOS version 7. This is a major update to the clusters and requires:

  • All Jobs to be terminated

  • All Queues to be empty

  • All Users to be logged out

Phase 2 (August 30, 2022, All day)

On the 30th, we will

  • Reinstall all nodes with EL8

  • Reinstall all rebuilt applications

  • Update Slurm to the latest production release

As reminder, a major OS update has significant impact, these include

  • Most user-compiled code will need to be rebuilt.

  • Some conda environments may need to be rebuilt to work properly with new system libraries.

  • A few older applications simply will not build on EL8 and will be removed.

  • The planned updates and sunset of RCIC-installed applications are available online.

FAQ

Is keeping CentOS7 an option?

Not really. It will be end-of-life in Jun 2024. We are already seeing commercial vendors ending support for CentOS7. The viability of CentOS7 as a functional OS will progressively diminish significantly over the next 12 months.

I really need some of the removed software, what do I do?

Please contact us via our ticketing system. In some cases, we may be able to build a Singularity container with the older applications and dependencies.

Can I have queued jobs during this downtime?

No. The Slurm upgrade requires there to be neither queued nor running jobs.

Can I continue using my conda environment after the upgrade?

It is very likely that you will need to rebuild your conda environment. Even if we build the same version of conda many underlying included packages will be different and of a newer version. This may or may not require your conda environment rebuild. You will have to test your environment and rebuild it if broken. We provide a guide Building and using conda environments