Research IT

provides research data and computing technologies, consulting, and community for the UC Berkeley campus. Our goal is to advance research through IT innovation.

Status and Service Updates

Mon, 5/4/2026: Savio System Status

Savio is fully operational and is open for general access

✅ Compute
✅ Storage
✅ Login nodes

Fri, 5/1/2026: Savio System Status

⚠️ Compute
✅ Storage
✅ Login nodes

We're seeing that 78% of the compute nodes are in an up/OK state, so many of the compute nodes still need to be brought back online after the maintenance period yesterday. Only about 21% of savio3_htc nodes are up/OK. We are continuing to work on this issue and will continue to share updates.

Fri, 5/1/2026: Savio Urgent Maintenance Security patch update COMPLETE

The emergency operating system security patch update has been successfully applied. The update has been deployed across the login, data-transfer, and compute nodes, and all updated nodes have been rebooted. You may now restart your jobs and resume normal workflows. Access to data and other services—including Open OnDemand and Globus—is available again. Thank you for your patience and cooperation. If you have any questions or concerns, please contact us.

Thurs, 4/30/2026: Savio Urgent Maintenance : Security patch update

We are writing to let you know that we need to perform urgent maintenance on the Savio Cluster to apply an immediate security patch to the operating system. To ensure the system is fully protected and to properly track patching completion, we decided to perform an emergency OS security patch update that requires a full system reboot across the login, data-transfer, and compute nodes. Impact: During the reboot and patching process, all running jobs will be canceled and will require resubmission. Access to the system through Open OnDemand will be unavailable, and file transfer via command line or Globus won't be possible. If your workflow supports checkpoint/restart, you may be able to reschedule from checkpoint files after the update. This patch is very crucial for the security and safety of the entire system. We appreciate your understanding and patience as we work through this. We plan to complete this work by tomorrow afternoon (Fri., 5/1/26).

Thurs, 4/30/2026: Savio System Status

Savio is fully operational and is open for general access

✅ Compute
✅ Storage
✅ Login nodes

Wed, 4/29/2026: Savio System Status

Savio is fully operational and is open for general access

✅ Compute
✅ Storage
✅ Login nodes

Tues, 4/28/2026: Savio System Status

Savio is fully operational and is open for general access

✅ Compute
✅ Storage
✅ Login nodes

Mon, 4/27/2026: Savio System Status

Savio is fully operational and is open for general access

✅ Compute
✅ Storage
✅ Login nodes

We have taken steps to restore connectivity to the Lustre storage system. We were able to restore one additional Lnet router, and we now have three out of four back online. Spot checks confirm that /global/scratch is mounting successfully on nodes that had previously failed to mount and were marked offline by Slurm. Please submit a ticket if you encounter any additional issues.

View all status and service updates

News Articles

Research IT is Hiring Graduate Student Domain Consultants

April 9, 2026

Research IT is hiring for multiple graduate student Domain Consultant positions. These are flexible, 10% to 25% (4-10 hours/week) remote appointments. We are seeking diverse and curious candidates from across scholarly disciplines who are enthusiastic about supporting campus researchers by...Read more about Research IT is Hiring Graduate Student Domain Consultants

Introduction to High Performance Computing with Savio Training

March 31, 2026

Tuesday, April 21, 2026, 3:40-5pm

Hybrid: in person at the Academic Innovation Studio (Dwinelle Hall) and via Zoom

Registration


Research IT is offering an...Read more about Introduction to High Performance Computing with Savio Training

Savio Computing Cluster Current Operational Status, Risk Mitigation, and Best Practices

February 24, 2026

We would like to provide an update on the Savio supercomputing cluster global scratch parallel file system service restoration and long-term stability. On Friday, January 23, 2026 the scratch file system on Savio became completely unavailable on the Savio cluster’s...Read more about Savio Computing Cluster Current Operational Status, Risk Mitigation, and Best Practices