Planned Outage of HPC ‘Deep Thought’ on 4 September 2019

On 4 September 2019 there will be a one-day outage of the Flinders high performance computing (HPC) facility Deep Thought, so that the HPC management tool SLURM can be enhanced.

The SLURM update will allow:
• Enhanced queue management
• More efficient use of available resources including specialist data nodes (GPU).

The following transition plan will affect Deep Thought users:
• 30 August 2019 — No new Deep Thought jobs can be submitted.
• 4 September 2019 — Deep Thought is taken offline for the day. The configuration of the SLURM tool is modified.

It is critical that all Deep Thought jobs finish before 4 September 2019.

Any jobs running on 4 September 2019 will be stopped, potentially resulting in lost data.
Users should update their SLURM scripts to take full advantage of the improvements.

Digital Research Services will provide further details to users via email in the week leading up to the change to assist them in making these changes.

If you have any questions, don’t hesitate to contact Digital Research Services at DigitalResearchServices@flinders.edu.au.

Thank you for your understanding, and we apologise for any inconvenience.

Posted in
DRS DRS Blog Post eResearch HPC

Leave a Reply