qcloud-user 1.0.15

Q-Cloud CLI for users

---

Q-Cloud User Documentation

---

Setup

Before submitting any calculations, you will need to install and configure the Q-Cloud command line interface:

python3 -m pip install  qcloud_user
qcloud --configure

You will be prompted for several configuration values that can be obtained from your Q-Cloud administrator. Alternatively, if your administrator has provided these details in a file, then you can provide the file name as an argument:

qcloud --configure user_info.txt

You should have received an email with an initial password for your account, and you will be prompted to change this the first time you attempt to submit a job.

Job Control

Submitting Jobs

Use the --submit option to submit Q-Chem jobs to the cluster, e.g.:

qcloud --submit job1.inp job2.inp [...]

Several jobs can be submitted at the same time and they will be submitted with the default queue parameters. If there are no compute nodes available, the jobs will sit in the QUEUED state for a couple of minutes while a fresh compute node is launched and configured. Once the queue has cleared, the compute nodes will automatically shut down after the configure time frame (default is five minutes).

Jobs will be submitted with the default queue parameters which are determined during the cluster setup (contact your QCloud administrator for details). Scratch space is set explicitly, memory is determined by the instance type selected and compute time is unlimited. If each job is run on a separate instance (by requesting all the instance cores) then these are the relevant default values.

If you want to override these values, or pass additional parameters to the SLUM scheduler, then you can add these to the first line of the input file as you would specify command line options to sbatch. For example: line of the Q-Chem input file. For example, the following limits the job to 1 hour and memory to 4G:

--time=1:00:00  --mem=4G
$molecule
0  1
he
$end
$rem
...

The number of threads can be specified by using the --ncpu flag, for example:

qcloud --submit --ncpu 4 job1.inp 

Note that if the number of threads specified exceeds the number of cores on an individual compute node, the job will not run. Your QCloud administrator will be able to inform you what this limit is.

If the job submission is successful, a unique job identifier will be returned:

[✓] Submitted job id gv6uqutvNmU0:             helium

A local registry of these IDs is kept, so it is not essential to use them in the commands below. However, they may be required to disambiguate multiple jobs submitted with the same input file basename.

Monitoring Jobs

To monitor the progress of jobs, use the --status option along with a string, which could be the file name, job ID or substring:

qcloud --status <jobid|jobname> 

The progress of jobs in the RUNNING state can be obtained using:

qcloud --tail <jobid|jobname> 

A job in the QUEUED or RUNNING state can be cancelled, which will remove it from the queue:

qcloud --cancel <jobid|jobname>

Downloading Results

Once a job in in the ARCHIVED state, it can be downloaded from the S3 bucket onto the local machine:

qcloud --get <pattern> 

The download will create a new directory with the same basename as the input file containing the output from the calculation.

Jobs in the DOWNLOADED state can be cleared from the job registry on the local machine:

qcloud --clear <pattern> 

Note that this does not remove the results from the S3 bucket. If you want to remove the job from the registry regardless of status, use the --remove option.

Other commands

The following will give a full list of commands available using the CLI:

qcloud --help

Troubleshooting

If you encounter additional problems not covered in this guide, please contact your Q-Cloud administrator or email [e-mail] for assistance.