Pseudo-terminal will not be allocated because stdin is not a terminal.
********************************************************************************
*  Welcome to                                                                  *
*       _ _   ___        _______ _     ____                                    *
*      | | | | \ \      / / ____| |   / ___|     Juelich Wizard                *
*   _  | | | | |\ \ /\ / /|  _| | |   \___ \        for                        *
*  | |_| | |_| | \ V  V / | |___| |___ ___) |      European Leadership         *
*   \___/ \___/   \_/\_/  |_____|_____|____/         Science                   *
*                                                                              *
********************************************************************************
* Information about the system, latest changes, user documentation and FAQs:   *
*  -> https://go.fzj.de/JUWELS   -> https://go.fzj.de/juwels-known-issues      *
* JUWELS cluster slurm job reports:                                            *
*  -> https://go.fzj.de/llview-juwels                                          *
********************************************************************************
* Status information also at https://go.fzj.de/status-juwels-cluster           *
********************************************************************************
                                                          2026-06-12T13:00+0200
Information - Slurm instabilities
This week there were 2 instances (one on Monday and one on Wednesday) where the
slurm controller failed and stopped responding to commands. Friday after 13:00
the rate of these failures increased. The root cause is unfortunately not clear,
but there are indications of a deadlock inside the controller code. No changes
have been done on the system in the last weeks that could have contributed to
this issue.

This implies that until this is solved issuing slurm commands might ocassionally
fail during some minutes, and that a system-wide reservation might be put in
place if necessary to debug further. Running jobs are not affected. We have put
measures in place to monitor the issue closer and have automatic recovery
actions outside of working hours.
--------------------------------------------------------------------------------
                                                          2026-07-05T00:30+0200
Critical Incident - Cooling failure in the datacenter
At 00:30 a general cooling failure on the datacenter resulted in an overheating
of all systems installed in the 16.4 building. Cooling seemed restored at 03:30.
The root cause is being investigated.

As a result JUWELS Cluster is at the moment completely unavailable. JUWELS
Booster and JURECA-DC are at the moment online, even though in the event JURECA-
DC lost 2 complete racks (a CPU-only rack and a GPU rack).

# Update 2026-07-05T08:15:00

JUWELS Booster and JURECA-DC are NOT online. The cooling seems stable but
production won't be resumed until it is confirmed that the cooling
infrastructure can handle the resulting load reliably.
--------------------------------------------------------------------------------
                                              Starting on 2026-07-07T09:00+0200
Planned - Software maintenance
On 2026-07-07 there is a maintenance for software updates on JUWELS. Login nodes
are updated, so a short interruption is to be expected. Compute nodes are
updated and full availability is expected during the second half of the day.
********************************************************************************
+------------------------------------------------------------------------------+
| This node is in maintenance. Overall system state visible on the status page |
+------------------------------------------------------------------------------+
Connection closed by 134.94.0.104 port 22