Scalabiliy of the shared disk size on a cluster

Question

I created the HPC cluster using parallel cluster 2.10. I would like to run cluster for long time and host my customers' simulations on it. My main requirement is that the HPC cluster is scalable, and the pricing is proportional to the amount of resources used.  
While this is the fact in case of compute nodes and they are added and removed whenever they are needed, I am not quite sure if the size of the shared EBS resources can be changed. I am confused, since the documentation of the parallel script says one cannot change `volume_size` after the cluster is created. On the other hand in the AWS console->EC2->EBS->Volumes there is an option to resize volumes. I am just wondering if this would work for the cluster. If not, what is the suggested solution to improve the cluster disk size scalability?

Answer

Hi @mfolusiak,  
  
"update not supported" policy for the volume_size parameter means that ParallelCluster is not able to perform all the required operations to update the size of the volume in the head node.  
  
Updating the file system after a resize operation requires commands specific to your operating system.  
  
You can anyway follow the general instructions to modify your volume size from AWS Console and manually change the size in the head node: https://docs.aws.amazon.com/AWSEC2/latest/UserGuide/requesting-ebs-volume-modifications.html#modify-ebs-volume  
  
Let us know if it helps.

Scalabiliy of the shared disk size on a cluster

関連するコンテンツ