[Lilab] Fwd: [UCR HPCC] Summer Maintenance Shutdown UPDATE 2

[Wei Vivian Li]

---------- Forwarded message ---------
From: HPCC_noreply <no-reply at hpcc.ucr.edu>
Date: Mon, Aug 4, 2025 at 1:39 PM
Subject: [UCR HPCC] Summer Maintenance Shutdown UPDATE 2
To: HPCC User <weil at ucr.edu>


[UCR HPCC] Summer Maintenance Shutdown UPDATE 2
[image: UCR High Performance Computing Center]
------------------------------

HPCC Summer Maintenance Shutdown UPDATE 2

When

   - Begins: Sunday, August 10, 9pm
   - Ends: Tuesday, August 12, evening
   - Follow-up outage: Early September (tentative)

Description

(Note: a few users received an email with "TEST1" in the subject. This
email is being resent with the correct subject line to all users.)

Dear HPCC Users,

We've just been informed that there was a miscommunication between UCR and
the electrical contractors regarding the exact date of the SOMED-1 power
shutdown. Instead of early Friday (8/8), power is scheduled to go down
early Monday (8/11).

In the interest of minimizing downtime for our users, we've decided once
again to shift our maintenance window. We will now plan to shutdown the
cluster on *Sunday, August 10, at 9pm*, and bring everything back up by
Tuesday evening.

HPCC apologizes for the uncertainty and disruptions caused by these sudden
developments. These maintenance events are necessary for the long-term
health of the cluster and related systems. Thanks for your continued
support and understanding.

Impact on Slurm Jobs, "ReqNodeNotAvail"

A Slurm reservation block has been set to prevent jobs from running during
the maintenance window. A previous email advised that some users may have
their jobs cancelled for the shutdown; this is no longer true based on the
new date and time.

If you have a job that extends into the maintenance window, it will stay in
the queue with the reason *"ReqNodeNotAvail, Reserved for maintenance"*. If
you want your job to start sooner, you will need to shorten the requested
time for your job such that it completes before maintenance begins
(approximately 6-7:30:00 as of this email). Otherwise, you may leave the
job as is, and it will start automatically once maintenance is done.

Other Services Impacted

Most services hosted by HPCC will be down during this period. This
includes, but is not limited to:

   - Cluster access (SSH, Slurm, data access), including web browser (HTTP)
   access to user data.
   - All services provided by OnDemand: RStudio/Posit, MATLAB, Jupyter, etc.

*Normal email support (support at hpcc.ucr.edu <support at hpcc.ucr.edu>) will be
down late Sunday to early Monday.* Our email support systems should come
back online around Monday noon. HPCC's Slack workspace (
https://ucr-hpcc.slack.com) will be the main means of group communication
with users during maintenance events. We strongly encourage users to become
familiar with our Slack workspace ahead of the shutdown.
------------------------------

Contacts

Austin Leong
Sr. HPC Systems Administrator
austin.leong at ucr.edu

Emerson Jacobson
HPC Systems Administrator
emerson.jacobson at ucr.edu

Thomas Girke
Professor
Director of HPC Center
thomas.girke at ucr.edu
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.ucr.edu/pipermail/lilab/attachments/20250804/94d574ec/attachment-0001.htm>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: noname
Type: image/png
Size: 19073 bytes
Desc: not available
URL: <https://lists.ucr.edu/pipermail/lilab/attachments/20250804/94d574ec/attachment-0001.png>