siqiliocean / htool

A collection of Bash scripts for managing tasks on a computing cluster managed by PBS, including checking node status, querying job information, and submitting jobs.

htool

htool is a collection of Bash scripts designed for managing tasks on a computing cluster managed by PBS. These scripts provide functionality to check node status, query job information, and submit jobs.

Scripts

hnode

hnode is a script to display information about nodes on a cluster, including all nodes, used nodes, rest nodes, and idle nodes of a specified queue.

hscan

hscan is a script to check the status of nodes in a cluster, determining if a node is online, idle, or occupied based on ping availability and CPU usage.

hstat

hstat is a script to query information about jobs running on a computing cluster using qstat, providing details such as job ID, queue name, processor information, node information, memory usage, and running path.

hscript

hscript is a script to copy the ${HTOOL}/hrun_exp to your run directory.

hrun_exp

hrun is an example script to submit jobs on the 'hydra' cluster using PBS directives. It allows users to customize job settings such as user name, job name, log file names, node number, processor per node, queue name, and executable command. It takes seconds to find out idle nodes.

Usage

1. Clone the 'htool' repository:

   git clone https://github.com/SiqiLiOcean/htool.git

2. Add the htool directory to PATH. For csh

   setenv HTOOL /YOUR_htool_PATH 
   setenv PATH ${HTOOL}:$PATH

   For bash

   export HTOOL /YOUR_htool_PATH 
   export PATH=${HTOOL}:$PATH

3. Add the execute permission

   chmod u+x ${HTOOL}/h*

4. Run the scripts as needed, following the usage instructions provided in each code.

Example

There is a 'hello' example in the directory exp for testing.

1. Go into the exp directory.

   cd exp

2. Compile the Fortran code.

   make

3. Copy hrun_exp here as hrun and edit it.

   hscript hrun

4. Submit the job in the parallel way.

   bash hrun

Author

htool is maintained by Siqi Li, SMAST.