Overview
The main point of using lonestar6 is that it is a massive computer cluster. If we run a command when logged into lonestar6, we are running it on one of the three low memory, low power "head" or "login" nodes on TACC. When we do serious computations that are going to take more than a few minutes or use a lot of RAM, we need to submit them to one of the other 560 computer nodes and 71, 680 processor cores on lonestar6.
...
Diagram of how a job gets run on lonestar6
Explanation
Start at the bottom - that's what you want: one of lonestar6 560 compute nodes running your specific program (bowtie mapping in this case).
...
Example: If you have two jobs in your commands file and you do not use wayness, by default, both jobs will be run on the same node and will be allocated 48gb/24 (~2gb) of memory per job (on lonestar6). If you use wayness of 1, each job will be run on a different node and each job will be allocated 24gb of memory.
...
**Wayness (commands/tasks per node)
Wayness sets how many commands/tasks are run on each compute node. By default, wayness will be 64 (equal to the number of physical cores per node on lonestar6). Each task will then get 1/64th of the memory available for one node (256 GB of memory) = 4 GB per task. Often, that is not enough memory per task or you may not even have 64 tasks in your commands file. Setting wayness to a number smaller than the default will allocate more memory per task.
tasks per node (wayness) | cores available to each task | memory available to each task |
|---|---|---|
| 1 | 64 | 256 GB |
| 2 | 32 | 128 GB |
| 4 | 16 | 64 GB |
| 8 | 8 | 32 GB |
| 16 | 4 | 16 GB |
| 64 | 1 | 4 GB |
| Code Block |
|---|
launcher_creator.py -n test -t 01:00:00 -j commands -q development -a OTH21164 |
...