CALCLOUD-462 Increase disk size for reprocessing MVMs

[sashawood]

Same as #219 plus:

1. black formatting
2. Better commenting of the connection between LENGTH_LADDER in deploy.sh and length(ladder) in locals.tf.

---

Previously, the terraform scripts:

created 1 launch template called hstsdp with a disk size of 150GB.
used a variable called local.ladder with 4 rungs - 02g, 08g, 16g, 64g
created 4 job queues, compute environments, and job definitions - one each for each of the 4 local.ladder rungs

I have modified the scripts to:

create 2 launch templates: hstdp["default"] with a disk size of 150GB and hstdp["large-volume"] with a disk size of 300GB.
use a local.ladder var with 8 rungs - 02g, 08g, 16g, 64g and 02g-large-volume, 08g-large-volume, 16g-large-volume, 64g-large-volume
create 8 job queues, compute environments, and job definitions - one each for each of the local.ladder rungs

In addition,

I modified multiple places where we explicitly used to iterate over 4 rungs of local.ladder to now use iterate over length(local.ladder) rungs.
In plan.py, I modified functions to use "default" AWS resources for ipppssoots and svms, and "large-volume" resources for mvms.
I updated the tests to include the new queues, compute environments, and definitions.

I have run this with tests 08106 and 08107 and observed the ipppssoots and svms are routed to default volume size EC2 instances, and mvms are routed to large-volume EC2 instances.
https://jira.stsci.edu/browse/CALCLOUD-462?focusedId=1037304&page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#comment-1037304

[sashawood]

Closing without merge, because this was created due to confusion.