the t2.micro instance has 1 GB RAM and 4 GB swap. it ran out of swap space. probably can't hold all of the TCGA datasets in memory. upgraded to t2.small that has 2 GB RAM. should consider doubling the swap space to 8 GB. could also split off the Rserver to a separate VM so it doesn't hog all the resources. additional Rserves would also distribute the memory load as each named dataset has only one cache resident amongst all the rserves.
Last edit: Keith Ching 2015-01-06
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
some of the larger analyses are sucking all the RAM and the whole 8GB of swap. will try to add another 8GB of swap to accomodate the paltry RAM of the AWS VM. The current production machine has 64GB of RAM and 500GB swap so I never noticed the problem.
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
the t2.micro instance has 1 GB RAM and 4 GB swap. it ran out of swap space. probably can't hold all of the TCGA datasets in memory. upgraded to t2.small that has 2 GB RAM. should consider doubling the swap space to 8 GB. could also split off the Rserver to a separate VM so it doesn't hog all the resources. additional Rserves would also distribute the memory load as each named dataset has only one cache resident amongst all the rserves.
Last edit: Keith Ching 2015-01-06
added an additional 4GB swap partition, for a total of 8GB swap on two separate SSDs.
some of the larger analyses are sucking all the RAM and the whole 8GB of swap. will try to add another 8GB of swap to accomodate the paltry RAM of the AWS VM. The current production machine has 64GB of RAM and 500GB swap so I never noticed the problem.
ok, 24GB of swap.. it'll probably get burned up..