Differenze

Queste sono le differenze tra la revisione selezionata e la versione attuale della pagina.

--- oph:cluster:messages [2026/05/05 07:50] – Avviso sovratemperatura diego.zuccato@unibo.it
+++ oph:cluster:messages [2026/06/11 09:47] (versione attuale) – [2026-06-11] è necessario richiedere le GPU perché il job possa usarle diego.zuccato@unibo.it
@@ Linea 15: / Linea 15: @@
 ===== 2026 =====
-==== 2026-04-05 ====
+==== 2026-06-11 ====
+Reconfiguring GPU nodes: when requesting a GPU node you have to *also* specify --gpus=N to have N GPUs assigned to your job. Other restrictions still apply, including allocation by socket (max 2 jobs per node).
+==== 2026-05-19 ====
+AC is now OK, the cluster have already been resumed.
+==== 2026-05-06 ====
+Started resuming some nodes. The biggest conditioner is still broken but the others have been fixed and are currently working. Hope not to have to shutdown again.
+==== 2026-05-05 ====
-<WRAP center round important>
 The server room is experiencing overtemperature due to a failed AC: many (not all) nodes are being drained and will be resumed ASAP.
-</WRAP>