Mild clarity issue in NVMe Usage Example
Closed this issue · 1 comments
There are some issues with clarity in the NVMe Usage Example section:
- there is a reference to test_nvme.sh, which is not used at all in the example batch script
- There is a Job submssion script: check_nvme.lsf. in bold in the middle of the paragraph (presumably this was meant to be the title of the code block below the paragraph)
- In the table comparing job scripts that use GPFS vs NVMe, there is no explanation of what
NAS-BTIO
,btio
, andinput*
refer to. I can understand based on context clues that maybeinput*
refers to files thatbtio
will write to, or something like that. But a small note explaining that might be helpful.
If you can provide some info on this, especially the third point, I can submit a pull request with these fixed.
These docs are pretty awesome otherwise!
-
I think the .sh should be .lsf
-
Yes, sometimes the text are merge from various other texts.
-
NAS BTIO was used for the example, you can find more information: http://www.softeng.rl.ac.uk/st/archive/SoftEng/SESP/Publications/paraio/paraio/node8.html and https://www.nas.nasa.gov/assets/pdf/techreports/2003/nas-03-002.pdf.
Something like:
BT-IO is part of the NAS parallel benchmark set, which is based on one of the computational kernels. The BT benchmark is based on a CFD code that uses an implicit algorithm to solve the 3D compressible Navier-Stokes equations. The benchmark requires an input file with the 3D dimensions as also the number of the timesteps for the solution.