This shows you the differences between two versions of the page.
Both sides previous revisionPrevious revisionNext revision | Previous revision | ||
start [2016/02/24 21:48] – neyron | start [2018/05/27 11:59] (current) – neyron | ||
---|---|---|---|
Line 14: | Line 14: | ||
{{ :: | {{ :: | ||
* Batch and Interactive jobs | * Batch and Interactive jobs | ||
+ | * Advance Reservation | ||
* Admission rules | * Admission rules | ||
* Walltime | * Walltime | ||
Line 23: | Line 24: | ||
* Check compute nodes before launching | * Check compute nodes before launching | ||
* Epilogue/ | * Epilogue/ | ||
- | * Activity | + | * Jobs and resources |
* No Daemon on compute nodes | * No Daemon on compute nodes | ||
- | * rsh and ssh as remote execution protocols (managed by Taktuk) | + | * SSH based remote execution protocols (managed by [[http:// |
* Dynamic insertion/ | * Dynamic insertion/ | ||
* Logging | * Logging | ||
* Backfilling | * Backfilling | ||
* First-Fit Scheduler with matching resource | * First-Fit Scheduler with matching resource | ||
- | * Advance Reservation | + | * On demand OS deployment |
- | * Environnement of Demand | + | * Grid computing support |
- | * Grid integration | + | |
- | * Simple Desktop Computing Mode | + | |
===== Why using OAR ===== | ===== Why using OAR ===== | ||
Line 90: | Line 89: | ||
* scalable exploiting tool Taktuk. | * scalable exploiting tool Taktuk. | ||
- | It is flexible enough to be suitable to manage | + | It is flexible enough to be suitable to manage |
- | OAR advantages: | + | Features: |
- | * No specific | + | * Only need a SSH daemon on nodes. |
- | * No dependence on specific | + | * No tied to any specific libraries |
- | * Upgrades are made on the servers, nothing | + | * Cpuset/ |
- | * CPUSET (2.6 linux kernel) integration which restricts | + | * Configurable for multicore, hyperthreading and GPUs |
- | * All administration tasks are performed with the taktuk command (a large scale remote execution deployment): | + | * Remote procedure calls using the [[http:// |
- | * Hierarchical resource requests (handle | + | * Hierarchical resource requests (multiple/heterogeneous clusters |
- | * Gantt scheduling (so you can visualize the internal scheduler decisions). | + | * Gantt scheduling (can visualize the internal scheduler decisions). |
* Full or partial time-sharing. | * Full or partial time-sharing. | ||
* Checkpoint/ | * Checkpoint/ | ||
- | * Licences | + | * Support for software licenses |
- | * Best effort jobs : if another job wants the same resources | + | * Best effort jobs: such a job is stopped automatically as soon as another job requires |
- | * Environment deployment support (Kadeploy): http:// | + | * Special job types: |
- | + | * deploy: support for in job OS deployment with software such as [[http:// | |
- | Other more common features: | + | * cosystem: support for a delegation to another job and resource management system |
- | + | * noop: reservation only jobs (no execution) | |
- | * Batch and Interactive jobs. | + | * placeholder, |
- | * Admission rules. | + | * Batch and Interactive jobs |
- | * Walltime. | + | * Advance reservations |
+ | * Admission rules | ||
+ | * Job walltime | ||
* Multi-schedulers support. | * Multi-schedulers support. | ||
- | * Multi-queues with priority. | + | * Multi-queues with priority |
- | * Backfilling. | + | * First-Fit Scheduler |
- | * First-Fit Scheduler. | + | * Moldable |
- | * Reservation. | + | |
- | * Support of moldable | + | |
- | * Check compute nodes. | + | |
* Epilogue/ | * Epilogue/ | ||
- | * Support of dynamic nodes. | + | * Dynamic resources definition |
* Logging/ | * Logging/ | ||
* Suspend/ | * Suspend/ | ||