Infrastructure renewal: Difference between revisions
No edit summary |
No edit summary |
||
| Line 6: | Line 6: | ||
<!--T:2--> | <!--T:2--> | ||
Our Advanced Research Computing infrastructure is undergoing major changes to always provide better High Performance Computing (HPC) and Cloud services for Canadian research. This page will be regularly updated to keep you informed of the activities concerning the transition to the new equipment.<br> | Our Advanced Research Computing infrastructure is undergoing major changes to always provide better High Performance Computing (HPC) and Cloud services for Canadian research. This page will be regularly updated to keep you informed of the activities concerning the transition to the new equipment.<br><br> | ||
The infrastructure renewal will replace the nearly 80% of our current equipment that is approaching end-of-life. The new equipment will offer faster processing speeds, greater storage capacity, and improved reliability. | The infrastructure renewal will replace the nearly 80% of our current equipment that is approaching end-of-life. The new equipment will offer faster processing speeds, greater storage capacity, and improved reliability. | ||
| Line 19: | Line 19: | ||
<!--T:4--> | <!--T:4--> | ||
=Technical specifications= | =Technical specifications= | ||
Technical specifications for each new system will be provided further down this page in future updates. Generally, they will be similar in architecture to the current systems, but with considerably increased capacity and performance.<br> | Technical specifications for each new system will be provided further down this page in future updates. Generally, they will be similar in architecture to the current systems, but with considerably increased capacity and performance.<br><br> | ||
For example, we expect to have fewer compute nodes, but each node will have a significant increase in the number of its cores, for an overall increase in the total number of CPU cores. | For example, we expect to have fewer compute nodes, but each node will have a significant increase in the number of its cores, for an overall increase in the total number of CPU cores. | ||
| Line 27: | Line 27: | ||
<!--T:6--> | <!--T:6--> | ||
==System outages== | ==System outages== | ||
An intense period of work will be conducted in the winter of 2024-2025 and spring of 2025. During the installation and the transition to the new systems, outages will be unavoidable due to constraints on space and electrical power. <br> | An intense period of work will be conducted in the winter of 2024-2025 and spring of 2025. During the installation and the transition to the new systems, outages will be unavoidable due to constraints on space and electrical power. <br><br> | ||
We recommend that you consider the possibility of outages when you plan research programs, graduate examinations, etc. | We recommend that you consider the possibility of outages when you plan research programs, graduate examinations, etc. | ||
<!--T:7--> | <!--T:7--> | ||
==Resource Allocation Competition (RAC)== | ==Resource Allocation Competition (RAC)== | ||
The Resource Allocation Competition will be impacted by this transition, but the application process remains the same. Application deadline this year is October 30, 2024.<br> | The Resource Allocation Competition will be impacted by this transition, but the application process remains the same. Application deadline this year is October 30, 2024.<br><br> | ||
2024/25 allocations will remain in effect on retiring clusters while each cluster remains in service. The 2025/26 allocations will be implemented everywhere once all new clusters are in service.<br> | 2024/25 allocations will remain in effect on retiring clusters while each cluster remains in service. The 2025/26 allocations will be implemented everywhere once all new clusters are in service.<br><br> | ||
Because the old clusters will mostly be out of service before all new ones are available, if you hold both a 2024 and a 2025 RAC award you will experience a period when neither award is available to you. You will be able to compute with your default allocation (<code>def-xxxxxx</code>) on each new cluster as soon as it goes into service, but the 2025 RAC allocations will only become available when all new clusters are in service. | Because the old clusters will mostly be out of service before all new ones are available, if you hold both a 2024 and a 2025 RAC award you will experience a period when neither award is available to you. You will be able to compute with your default allocation (<code>def-xxxxxx</code>) on each new cluster as soon as it goes into service, but the 2025 RAC allocations will only become available when all new clusters are in service. | ||
Revision as of 16:18, 20 September 2024
Major upgrade of our Advanced Research Computing infrastructure[edit]
Our Advanced Research Computing infrastructure is undergoing major changes to always provide better High Performance Computing (HPC) and Cloud services for Canadian research. This page will be regularly updated to keep you informed of the activities concerning the transition to the new equipment.
The infrastructure renewal will replace the nearly 80% of our current equipment that is approaching end-of-life. The new equipment will offer faster processing speeds, greater storage capacity, and improved reliability.
The systems involved are
- Arbutus, cloud
- Béluga, compute cluster only (not cloud)
- Cedar, compute cluster and cloud
- Graham, compute cluster and cloud
- Niagara, compute cluster
Technical specifications[edit]
Technical specifications for each new system will be provided further down this page in future updates. Generally, they will be similar in architecture to the current systems, but with considerably increased capacity and performance.
For example, we expect to have fewer compute nodes, but each node will have a significant increase in the number of its cores, for an overall increase in the total number of CPU cores.
Impacts[edit]
System outages[edit]
An intense period of work will be conducted in the winter of 2024-2025 and spring of 2025. During the installation and the transition to the new systems, outages will be unavoidable due to constraints on space and electrical power.
We recommend that you consider the possibility of outages when you plan research programs, graduate examinations, etc.
Resource Allocation Competition (RAC)[edit]
The Resource Allocation Competition will be impacted by this transition, but the application process remains the same. Application deadline this year is October 30, 2024.
2024/25 allocations will remain in effect on retiring clusters while each cluster remains in service. The 2025/26 allocations will be implemented everywhere once all new clusters are in service.
Because the old clusters will mostly be out of service before all new ones are available, if you hold both a 2024 and a 2025 RAC award you will experience a period when neither award is available to you. You will be able to compute with your default allocation (def-xxxxxx) on each new cluster as soon as it goes into service, but the 2025 RAC allocations will only become available when all new clusters are in service.
Status updates[edit]
For current outages please see the system status page.
Resource Allocation Competition (RAC)[edit]
| Sep 13, 2024 | The RFP processes for all sites except for Rorqual (the replacement of Béluga) have been completed, and purchase orders to vendors have been sent. The Rorqual storage RFP is still open and is scheduled to complete on Sep 18.
All sites are working on infrastructure (power and cooling) design and implementation. We are expecting some outages over the autumn for cabling and plumbing upgrades, and will update this page when we know more. |
| Sep 3, 2024 | Currently all sites are completing their Requests for Proposals, and have been working with the vendors on deliverables and purchase orders. |
Technical specifications[edit]
The sites cannot yet provide detailed technical specifications of the new systems. Generally, the new systems will be similar in architecture to the old systems but with considerably increased capacity and performance. For instance, we expect to have fewer compute nodes, but each node will have a significant increase in the number of cores due to the increase in the size of multi-core CPUs since 2017.
Resource Allocation Competition and renewals[edit]
The Resource Allocation Competition (RAC) and RAC renewals will be affected by this transition, but we are not changing the normal RAC process. Expect to see the usual announcements for the competition in September 2024. We expect to implement the 2025/26 allocations on the new machines when they become available so there may be some delay in RAC implementation. See RAC documentation available here.
System-specific updates[edit]
Arbutus[edit]
Coming soon
Béluga[edit]
The cluster that is being deployed to replace Béluga will be named Rorqual.
More details coming soon.
Cedar and Cedar Cloud[edit]
Coming soon
Graham and Graham Cloud[edit]
Coming soon
Niagara[edit]
Coming soon
Frequently asked questions[edit]
As we work on finalizing the details, here are a few key points to keep in mind.
|
|
Will data be copied to the new systems?[edit]
Data migration to the new systems is a site responsibility. Each site will let you know what you need to do and what will be done for you once the details are finalized.
When will outages occur?[edit]
Each site will have their own schedule for outages as the new equipment is installed and transitioned. Specific outages will as usual be described on the status pages (https://status.alliancecan.ca). We will also provide more general updates through this wiki page as we know more, probably in early autumn 2024. We will also periodically send emails with updates and outage notices.
Who should I contact for questions about the transition?[edit]
Contact our Technical support, but don't expect them to know a great deal more than you read here.
Will my jobs/applications run without change on the new system?[edit]
Generally yes, but with new CPUs and GPUs some codes may need recompiling or reconfiguring. More details will be provided during the transition.
Will the software from the old systems still be available?[edit]
Yes, our standard software environment will be available on the new systems.
Will there be staggered outages?[edit]
We will do our best to limit overlapping outages, but we are very constrained by delivery schedules and funding deadlines so there will probably be periods when many of our systems are simultaneously out. We’ll communicate all outages as early as possible.