September 2026

Date: Monday, September 28, 2026 (08:00-10:00)

Who is affected:
All users of the CS Department Beowulf high performance computing clusters,
known as ionic and neuronic.

All users of the CS Staff-managed public login systems, including the
cycles, courselab, and armlab systems.

What is happening:
CS Staff will apply emergency RHEL 9 security updates to the Ionic and
Neuronic clusters, as well as the Cycles, Courselab, and Armlab systems, to
address recently identified security vulnerabilities.

The Neuronic login node, Cycles, Courselab, and Armlab systems will be
updated and rebooted during the scheduled maintenance window. Users may
experience temporary service interruptions and will be unable to access the
affected systems while they are being rebooted.

Ionic and Neuronic compute nodes will be updated and rebooted on a rolling
basis as they become available, either after running jobs have completed or
when the nodes have been drained. This approach will allow us to perform
the necessary security updates while minimizing disruption to running
jobs.

Please note that we expect to complete the updates to the login systems
earlier than the scheduled maintenance window, but the extended time frame
allows for any unexpected issues that may arise.

Why is it happening:
This emergency maintenance is necessary to address security vulnerabilities
and protect the security and integrity of our computing infrastructure.

Applying these security updates promptly helps safeguard user data, reduce
the risk of unauthorized access, and maintain a secure and reliable
computing environment for all users.

We will post updates to the status page: www.csstaff.org
as necessary.

If this downtime will cause you undue hardship, please contact
csstaff@cs.princeton.edu immediately, so we can discuss options to reduce
any negative impact. Your patience is appreciated.

Sincerely,
CS Staff

Date: Tuesday, October 20, 2026 (06:00-18:00)

Who is affected:
ALL users of CS Department (or University) cluster computing systems
AND
ALL users of research rack space at the HPCRC (151 Forrestal Road)

What is happening:
On October 20, 2026, all “B”-side power at the HPCRC will be turned off and
unavailable for the full business day while preventative major maintenance
is performed on the building’s power systems. “B”-side power accounts for
half the power to the building, and ALL of the power to some racks occupied
by the Computer Science Department.

Other than the ionic and neuronic clusters, most research racks in the CS
portion of the HPCRC use “A”-side power. Nevertheless, it would be prudent
to anticipate that research racks may lose power at 07:00 and not regain
power before 15:00. It is recommended that researchers arrange to power off
their equipment prior to the 07:00 start time and be prepared to check its
health and restore services after the 15:00 end. CS Staff is announcing a
wider 12-hour window to allow for this time to orderly shut down clusters,
and to power back on after the work.

ALL ionic and neuronic cluster resources will be offline and unavailable.
ALL courselab and armlab systems will be offline and unavailable.
The “cycles” systems will be rebooted at the start of the window, but are
anticipated to be available.
All other central CS services (e.g. email, web service, storage) are
anticipated to remain available.

OIT’s announcement of this maintenance can be found here:
oit.princeton.edu/events/hpcrc-planned-downtime-maintenance

Why is it happening:
The power equipment at the HPCRC is overdue for major maintenance to reduce
the risk of unplanned failures and outages. This maintenance window, like
the earlier outage on June 9, allows for this maintenance to be performed
safely.

We will post updates to the status page: www.csstaff.org
as necessary.

If this downtime will cause you undue hardship, please contact
csstaff@cs.princeton.edu immediately, so we can discuss options to reduce
any negative impact. Your patience is appreciated.

Sincerely,
CS Staff