Add Lmod startup hook that prints an error when loading removed/relocated modules#962
Conversation
|
Instance
|
|
Instance
|
|
Instance
|
|
I tested this locally and it works, also when the module have been removed. The one gotcha I can think of is that the hook will trigger on all |
…_MODULES_CHECK env var
|
bot: build repo:eessi.io-2023.06-software instance:eessi-bot-mc-aws arch:x86_64/amd/zen2 |
Updates by the bot instance
|
Updates by the bot instance
|
Updates by the bot instance
|
|
New job on instance
|
|
New job on instance
|
|
New job on instance
|
|
New job on instance
|
|
New job on instance
|
|
New job on instance
|
|
New job on instance
|
|
New job on instance
|
|
New job on instance
|
|
New job on instance
|
|
Label |
|
@ocaisa deployed, so should be merged? |
|
I usually wait to see all the staging PRs ingested before merging (and that usually means I go away and do something else and then come back to it). Do others merge directly after merging the staging PRs (PRs like these don't get held back by missing software CI)? |
|
bot: build repo:eessi.io-2023.06-software instance:eessi-bot-mc-aws arch:x86_64/amd/zen2 |
Updates by the bot instance
|
Updates by the bot instance
|
Updates by the bot instance
|
|
New job on instance
|
|
New job on instance
|
|
New job on instance
|
|
New job on instance
|
|
New job on instance
|
|
New job on instance
|
|
New job on instance
|
|
New job on instance
|
|
New job on instance
|
|
New job on instance
|
This type of PR should get merged quickly, otherwise ongoing builds pick up on the changed file, and then it gets messy fast... |
|
All staging PRs merged |
|
PR merged! Moved |
|
PR merged! Moved |
Implements a fix for https://gitlab.com/eessi/support/-/issues/128, allowing us to remove the CUDA installations from the CPU prefixes.
Tested this a bit in the EESSI container on an AWS VM.
No accelerator stack in
$MODULEPATH:Now we add the accelerator prefix to
$MODULEPATH:Perhaps someone else could also take this for a spin?