-----------------------------------------------------------------------
<h1 align="center"> Q-Cloud Administrator Documentation </h1>
-----------------------------------------------------------------------
## Introduction
Q-Cloud is a framework that allows users to easily launch elastic compute
clusters on AWS and submit Q-Chem calculations to them. The clusters will
automatically expand with demand, up to the maximum size determined by
the type of Q-Cloud license purchased, and idle nodes will shut down to
minimize running costs.
Q-Cloud clusters are built around three separate AWS service stacks with the
following names and purposes:
1. _qcloud-users_: This stack controls user access to the cluster. It
launches a Cognito service which manages users, passwords and access tokens.
2. _qcloud-api-gateway_: This stack provides the REST endpoints for submitting jobs to
the cluster and for accessing the results from calculations.
3. _qcloud-cluster_: This is the actual compute cluster and consists of a head node
(which can be just a t2.micro instance) running a SLURM workload manager.
The head node is responsible for launching the compute instances which run
the Q-Chem calculations. The head node runs for the lifetime of the
cluster, but the compute nodes run on-demand and automatically terminate
when there are no jobs in the queue.
**Note:** Charges apply to many AWS services. See the Costs section for
further details and information on how to minimise the running costs of the
cluster.
## Prerequisites
The cluster administrator will need to have a valid AWS account and be able to
log into their account via the [console](https://signin.aws.amazon.com/) in
order to obtain access and secret access keys.
Users of the cluster do not require AWS accounts as their access to the
cluster is configured separately, (see the Adding users section below for
further details on setting up user accounts).
The Q-Cloud packages require you to have python (v3.9 or later)
installed on the local machine used to administer the cluster.
The following steps are required to set up a Q-Cloud cluster.
## 1) Obtain access keys
This step can be skipped if you have already configured the AWS CLI.
Open up the AWS [console](https://signin.aws.amazon.com/) in a browser. Search
for IAM and select the first hit. This brings up the IAM dashboard.
1. From the left hand panel select 'Users' and then click on the name of a user
with administrator privileges (i.e. the AdministratorAccess permissions
policy has been attached). Note this administrator user is only required
to create the qcloud_admin user with restricted permissions and is not used
to launch any of the Q-Cloud resources.
2. Select the 'Security credentials' tab and scroll down to the 'Access keys'
section and click the 'Create access key' button.
3. Select the 'Command Line Interface CLI' option as the use case, check
the confirmation checkbox and click the Next button.
4. Enter a description for the key and click 'Create access key'.
5. Make a note of both the access key and the secret access key. You will not
have another opportunity to see the secret key, so if you misplace it you
will need to deactivate the key and generate a new one. The access keys
look something like the following:
```
aws_access_key_id = AKIBS7OH2PUVOGH8A9ZT
aws_secret_access_key = 5rV7LzKMLKNUO+IupIgsdj876FSx8nPLdfy7F9A4
```
## 2) Install the qcloud_setup script
Download the install_qcloud.sh script, add executable permissions and run it:
```
wget https://downloads.q-chem.com/install_qcloud.sh
chmod +x install_qcloud.sh
./install_qcloud.sh
```
This will create a virtual environment with the required Python packages installed.
Make a note of the directory created for the virtual environment (this is output
at the end of the install script). To activate the virtual environment and run
the qcloud_setup script, simply type.
```
bash
cd /path/to/qcloud_venv
source env/bin/activate
qcloud_setup [options]
```
Note the bash line is only required if bash is not your regular shell.
To exit the virtual environment after setting up your cluster simply, type
```
deactivate
```
You will need to activate the virtual environment before using the qcloud_setup
command.
## 3) Configure cluster
Within the virtual environment, cluster configuration files can be generated with the command:
```
qcloud_setup --configure
```
This takes the user through an interactive setup process which produces a
configuration file with the name *qcloud-cluster.config* which can be viewed
before launching the cluster. Do not modify the contents of the configuration
file as it may cause subsequent steps in the setup to fail.
By default the name of the cluster is 'qcloud', but this can be changed using
the `--name <cluster_name>` option. This allows for multiple clusters to run
at the same time, if required. Note that if a non-default name is specified
then this name must be passed using the `--name` option to subsequent commands
(e.g. `--launch`) in order that they operate on the correct cluster.
The options specified in the configuration process are discussed in detail below.
#### 3.1) AWS Credentials
These should have been configured previously for the qcloud_admin user and, if
so, are automatically detected.
A new SSH key pair will be created to connect to the head node instance and the
private key will be saved in the qcloud\_\<region\>\_keypair.pem file. This
file needs to be available when performing many of the qcloud_setup commands,
either in the current directory or by copying it to the location of your other
key files (e.g. the ~/.ssh directory). Do not lose this private key as you
will be unable to connect to the head node without it.
#### 3.2) VPC setup
The Q-Cloud cluster runs inside a Virtual Private Cloud (VPC) and it is
recommended that a new VPC be created for running the Q-Cloud cluster.
However, because there is a limit on the number of VPCs per AWS account, this may
not be possible for some users. Please select whichever of the following
options is best for your case:
- Default: This option will use your default VPC and both the head node and
compute clusters will be in a public subnet (not recommended).
- Use existing: If available. The setup script will look for a previously
configured Q-Cloud VPC. This will have both public (for the head node)
and private (compute nodes) subnets configured.
- Create new: This will create a new VPC with public and private subnets.
The availability zone of the subnets does not matter.
#### 3.3) Compute instance types
The costs shown are indicative only and are for running each instance. They
do not include storage or network costs. Note the difference between c5 and
c5d instance types is the latter uses local NVMe-based SSD block storage which
is physically connected to the instance and therefore faster.
#### 3.4) Maximum compute nodes
This determines the maximum number of compute instances that can run concurrently.
If more jobs are submitted than this value, they will be queued until the resources
become available.
This value should be set to the same number as the number of seats purchased as
part of your license.
#### 3.5) Job files
This is the space allocated for the jobs volume which is attached to the head
node and shared between compute instances. This volume is used for output
files only and can be quite small. Output files are automatically deleted
after they have been copied to the S3 bucket.
#### 3.6) Scratch files
This determines the scratch volume size per instance. As an indicative cost,
100Gb costs around $10/month.
## 4) Launching the cluster
Once a configuration file has been generated, the service stacks can be
launched with the following:
```
qcloud_setup --launch
```
The configuration file is automatically updated with values for the place holders
as the various resources are created.
**Notes:**
1. Some steps in the launch process can take several minutes to complete, in
particular building the cluster stack. Migrating the Q-Cloud software to
your region can take a variable amount of time depending on network load.
2. If you terminate the the qcloud\_setup script during the creation process
the stack will continue to be created. Use the `--delete` option to
actually delete or stop the stack creation.
3. Interrupting the launch process may leave a temporary snapshot lying around.
To delete this, log into the [AWS console](https://signin.aws.amazon.com/) and
navigate to EC2 → Snapshots. Any Q-Cloud snapshots can be safely deleted
as long as you are not currently launching a cluster.
Once launched, you will need to send the following information to
license@q-chem.com to obtain your license activation key:
- Order number
- Name
- University / Institution
- Elastic IP address (provided as output from the launch command)
Once you have your activation key, you can install it as follows:
```
qcloud_setup --activation-key XXXX-XXXX-XXXX-XXXX-XXXX
```
This command requires access to the SSH key generated during the configuration
step, which should be either in the current directory or in your home directory
under ~/.ssh.
## Adding users
Before submitting jobs, a user will need to be added to the Cognito user pool:
```
qcloud_setup --add-user <user_name> --email <email_address>
```
Alternatively, you can add multiple users from a file:
```
qcloud_setup --add-users <file_name>
```
where the file\_name consists of a list of user names and email addresses with the
format:
```
elmo elmo@gmail.com
bigbird bigbird@gmail.com
...
```
A message will be sent to the user's email address with their user name and a
temporary password, which will need to be changed when first attempting to
submit a job.
Note that the number of users able to be added each day is limited to 50 due to
email limits in the Cognito service. If you need to add more than this you
will need to configure Cognito with the Simple Email Service' (SES) and add a
validated administrator email.
The cluster administrator will need to provide the server details to each
user which can be obtained by running the command:
```
qcloud_setup --user-info
AwsRegion us-east-1
CognitoUserPoolId us-east-1_KbkdtpKpW
CognitoAppClientId 2mgcn0o8fkakboq7jnqs6bd6ee
ApiGatewayId fkkxolpuo4
```
If these values are saved to a file:
```
qcloud_setup --user-info > user_info.txt
```
then the cluster users can use this file to configure the client:
```
qcloud --configure user_info.txt
```
Cluster users will need to install and configure the command line interface (CLI):
```
python3 -m pip install qcloud_user
```
this will install the qcloud command into their python environment and this can be configured
by running the following command and entering the values from the
`qcloud_setup --user-info` command
```
qcloud --configure
```
The first time a user interacts with the cluster they will need to enter the temporary
password emailed to them before resetting their password.
See the README.md file distributed with the qcloud\_user module for further
details on interacting with the server.
## Suspending a Cluster
It is possible to shut down the cluster head node and restart it at a later
time in order to minimise the running costs. If you plan to restart the
cluster, make sure to keep copies of the configuration file, ssh key (.pem
file) and activation key.
Use the `--suspend` option to delete only the cluster stack. You will be prompted
if you want to delete the stack, type 'y' to confirm.
```
qcloud_setup --suspend
Delete stack qcloud-cluster? [y/N] y
```
This terminates the head node, but leaves the API gateway and Cognito stacks
running, using minimal resources. Once the cluster has been deleted, the
results of previous jobs can still be accessed via the API gateway as these are
archived in an S3 bucket. You will, of course, be unable to submit further
jobs until the cluster has been restarted.
To re-launch the cluster, issue the following command in the same directory as
the configuration and license files:
```
qcloud_setup --launch
qcloud_setup --activation-key XXXX-XXXX-XXXX-XXXX-XXXX
```
**Note:** In order to restart a cluster, you must have configured the cluster to
use an elastic IP (EIP) address allocated to your AWS account. This EIP must be
available for reuse when re-launching the cluster, otherwise the Q-Chem license
will no longer be valid.
If you need to update the EIP of your host, please contact our support team at
support@q-chem.com for assistance.
## Teminating a cluster
Before terminating a cluster, ensure there are no jobs running in the queue and
that you have downloaded the results of any calculations you wish to keep. The
following command will delete the cluster:
```
qcloud_setup --delete
```
You will be asked whether you want to delete each of the 3 stacks,
and you should type 'y' for each.
To clean up all the resources allocated to the cluster, be sure to release the
elastic IP address and clean out any files created in the S3 bucket.
## Other options
Additional options for the setup script can be printed via:
```
qcloud_setup --help
qcloud_setup --info
qcloud_setup --list
```
You can also open an ssh connection to the head node using the following:
```
qcloud_setup --shell
```
## Costs
The costs associated with running the cluster hardware will be dominated by the
compute nodes used to run calculations. These costs depend on the type of node
and the region configured during setup and can be estimated using the
[AWS Cost Estimator](https://calculator.aws/#/addService/ec2-enhancement).
In addition to the compute-node costs, there are overhead costs associated with
running the head-node used for job submissions which will be incurred even if
there are no running jobs. A low-cost T2 instance is used for this purpose and
will attract a monthly cost of around $10, depending on the amount of job
storage selected at setup time.
The costs associated with the head-node can be minimized by suspending the
cluster when not in use. If this is done, there will still be a residual
charge for maintaining the elastic IP which is required for licensing. This
cost is approximately $3.65 per month.
*Note*: These costs are managed by AWS and will be charged to the account used
to launch the Q-Cloud cluster, they are separate from the Q-Cloud subscription
fee which is charged by Q-Chem Inc.
To monitor the costs associated with the cluster, the following will
give a summary of the costs recorded for the last 28 days:
```
qcloud_setup --costs
```
You can also set notifications for when expenditure reaches 75% and 99% of
a specified budget, or when the expenditure is predicted to exceed the
budgeted figure:
```
qcloud_setup --set-budget n
```
where n is a integer number of USD.
*Note*: Currently these cost settings apply to the entire AWS account.
## Troubleshooting
- When launching the cluster, if you receive permission problems associated
with lambda functions, check to ensure that the IAM policy (QCloudIamPolicy) has been
created in the same region as the cluster.
- When submitting jobs, a job ID should be returned. This is a random
sequence of 12 alphanumeric characters associated with your job. If you do
not see a job ID, it is possible the submit function has not been updated
with the instance ID of the head node. Re-run the launch command to trigger
this update:
```
qcloud_setup --launch
```
- If you encounter any problems not covered here, please contact our support team at
support@q-chem.com for assistance.
Raw data
{
"_id": null,
"home_page": "https://q-chem.com",
"name": "qcloud-setup",
"maintainer": null,
"docs_url": null,
"requires_python": ">=3.9",
"maintainer_email": null,
"keywords": "qcloud, qchem, q-cloud, q-chem",
"author": "Andrew Gilbert",
"author_email": "\"Q-Chem Inc.\" <support@q-chem.com>",
"download_url": "https://files.pythonhosted.org/packages/c5/dc/088b1229469bf8c0defee84cb1515a1048d27379a89268e64b8ff244d8c2/qcloud_setup-1.0.33.tar.gz",
"platform": null,
"description": "-----------------------------------------------------------------------\n\n<h1 align=\"center\"> Q-Cloud Administrator Documentation </h1>\n\n-----------------------------------------------------------------------\n\n\n## Introduction\n\nQ-Cloud is a framework that allows users to easily launch elastic compute\nclusters on AWS and submit Q-Chem calculations to them. The clusters will\nautomatically expand with demand, up to the maximum size determined by\nthe type of Q-Cloud license purchased, and idle nodes will shut down to \nminimize running costs.\n\nQ-Cloud clusters are built around three separate AWS service stacks with the\nfollowing names and purposes:\n1. _qcloud-users_: This stack controls user access to the cluster. It\n launches a Cognito service which manages users, passwords and access tokens.\n2. _qcloud-api-gateway_: This stack provides the REST endpoints for submitting jobs to\n the cluster and for accessing the results from calculations. \n3. _qcloud-cluster_: This is the actual compute cluster and consists of a head node\n (which can be just a t2.micro instance) running a SLURM workload manager.\n The head node is responsible for launching the compute instances which run\n the Q-Chem calculations. The head node runs for the lifetime of the\n cluster, but the compute nodes run on-demand and automatically terminate\n when there are no jobs in the queue.\n\n**Note:** Charges apply to many AWS services. See the Costs section for\nfurther details and information on how to minimise the running costs of the\ncluster.\n\n\n## Prerequisites\n\nThe cluster administrator will need to have a valid AWS account and be able to\nlog into their account via the [console](https://signin.aws.amazon.com/) in\norder to obtain access and secret access keys.\n\nUsers of the cluster do not require AWS accounts as their access to the\ncluster is configured separately, (see the Adding users section below for\nfurther details on setting up user accounts).\n\nThe Q-Cloud packages require you to have python (v3.9 or later)\ninstalled on the local machine used to administer the cluster. \n\nThe following steps are required to set up a Q-Cloud cluster.\n\n## 1) Obtain access keys\n\nThis step can be skipped if you have already configured the AWS CLI. \n\nOpen up the AWS [console](https://signin.aws.amazon.com/) in a browser. Search \nfor IAM and select the first hit. This brings up the IAM dashboard.\n\n1. From the left hand panel select 'Users' and then click on the name of a user\n with administrator privileges (i.e. the AdministratorAccess permissions\n policy has been attached). Note this administrator user is only required\n to create the qcloud_admin user with restricted permissions and is not used\n to launch any of the Q-Cloud resources.\n2. Select the 'Security credentials' tab and scroll down to the 'Access keys'\n section and click the 'Create access key' button.\n3. Select the 'Command Line Interface CLI' option as the use case, check\n the confirmation checkbox and click the Next button.\n4. Enter a description for the key and click 'Create access key'.\n5. Make a note of both the access key and the secret access key. You will not\n have another opportunity to see the secret key, so if you misplace it you\n will need to deactivate the key and generate a new one. The access keys \n look something like the following:\n```\naws_access_key_id = AKIBS7OH2PUVOGH8A9ZT\naws_secret_access_key = 5rV7LzKMLKNUO+IupIgsdj876FSx8nPLdfy7F9A4\n```\n\n\n## 2) Install the qcloud_setup script\n\nDownload the install_qcloud.sh script, add executable permissions and run it:\n```\nwget https://downloads.q-chem.com/install_qcloud.sh\nchmod +x install_qcloud.sh\n./install_qcloud.sh\n```\nThis will create a virtual environment with the required Python packages installed.\nMake a note of the directory created for the virtual environment (this is output\nat the end of the install script). To activate the virtual environment and run\nthe qcloud_setup script, simply type.\n```\nbash\ncd /path/to/qcloud_venv\nsource env/bin/activate\nqcloud_setup [options]\n```\nNote the bash line is only required if bash is not your regular shell.\nTo exit the virtual environment after setting up your cluster simply, type \n```\ndeactivate\n```\nYou will need to activate the virtual environment before using the qcloud_setup\ncommand.\n\n\n## 3) Configure cluster\nWithin the virtual environment, cluster configuration files can be generated with the command: \n```\nqcloud_setup --configure\n```\nThis takes the user through an interactive setup process which produces a\nconfiguration file with the name *qcloud-cluster.config* which can be viewed\nbefore launching the cluster. Do not modify the contents of the configuration\nfile as it may cause subsequent steps in the setup to fail.\n\nBy default the name of the cluster is 'qcloud', but this can be changed using\nthe `--name <cluster_name>` option. This allows for multiple clusters to run\nat the same time, if required. Note that if a non-default name is specified\nthen this name must be passed using the `--name` option to subsequent commands\n(e.g. `--launch`) in order that they operate on the correct cluster.\n\nThe options specified in the configuration process are discussed in detail below.\n\n\n#### 3.1) AWS Credentials\n\nThese should have been configured previously for the qcloud_admin user and, if\nso, are automatically detected. \n\nA new SSH key pair will be created to connect to the head node instance and the\nprivate key will be saved in the qcloud\\_\\<region\\>\\_keypair.pem file. This\nfile needs to be available when performing many of the qcloud_setup commands,\neither in the current directory or by copying it to the location of your other\nkey files (e.g. the ~/.ssh directory). Do not lose this private key as you\nwill be unable to connect to the head node without it.\n\n\n#### 3.2) VPC setup\n\nThe Q-Cloud cluster runs inside a Virtual Private Cloud (VPC) and it is\nrecommended that a new VPC be created for running the Q-Cloud cluster.\nHowever, because there is a limit on the number of VPCs per AWS account, this may\nnot be possible for some users. Please select whichever of the following \noptions is best for your case:\n\n- Default: This option will use your default VPC and both the head node and\n compute clusters will be in a public subnet (not recommended).\n- Use existing: If available. The setup script will look for a previously\n configured Q-Cloud VPC. This will have both public (for the head node) \n and private (compute nodes) subnets configured.\n- Create new: This will create a new VPC with public and private subnets. \n The availability zone of the subnets does not matter.\n\n\n#### 3.3) Compute instance types\n\nThe costs shown are indicative only and are for running each instance. They\ndo not include storage or network costs. Note the difference between c5 and\nc5d instance types is the latter uses local NVMe-based SSD block storage which\nis physically connected to the instance and therefore faster.\n\n\n#### 3.4) Maximum compute nodes\nThis determines the maximum number of compute instances that can run concurrently. \nIf more jobs are submitted than this value, they will be queued until the resources\nbecome available.\n\nThis value should be set to the same number as the number of seats purchased as\npart of your license.\n\n\n#### 3.5) Job files\nThis is the space allocated for the jobs volume which is attached to the head\nnode and shared between compute instances. This volume is used for output\nfiles only and can be quite small. Output files are automatically deleted\nafter they have been copied to the S3 bucket.\n\n\n#### 3.6) Scratch files\n\nThis determines the scratch volume size per instance. As an indicative cost,\n100Gb costs around $10/month.\n\n\n## 4) Launching the cluster\n\nOnce a configuration file has been generated, the service stacks can be\nlaunched with the following:\n```\nqcloud_setup --launch \n```\nThe configuration file is automatically updated with values for the place holders\nas the various resources are created. \n\n**Notes:** \n1. Some steps in the launch process can take several minutes to complete, in\n particular building the cluster stack. Migrating the Q-Cloud software to\n your region can take a variable amount of time depending on network load.\n2. If you terminate the the qcloud\\_setup script during the creation process\n the stack will continue to be created. Use the `--delete` option to\n actually delete or stop the stack creation. \n3. Interrupting the launch process may leave a temporary snapshot lying around.\n To delete this, log into the [AWS console](https://signin.aws.amazon.com/) and\n navigate to EC2 → Snapshots. Any Q-Cloud snapshots can be safely deleted\n as long as you are not currently launching a cluster.\n\nOnce launched, you will need to send the following information to\nlicense@q-chem.com to obtain your license activation key:\n\n- Order number\n- Name \n- University / Institution\n- Elastic IP address (provided as output from the launch command)\n\nOnce you have your activation key, you can install it as follows:\n```\nqcloud_setup --activation-key XXXX-XXXX-XXXX-XXXX-XXXX\n```\nThis command requires access to the SSH key generated during the configuration\nstep, which should be either in the current directory or in your home directory\nunder ~/.ssh.\n\n\n\n## Adding users\nBefore submitting jobs, a user will need to be added to the Cognito user pool:\n```\nqcloud_setup --add-user <user_name> --email <email_address>\n```\nAlternatively, you can add multiple users from a file:\n```\nqcloud_setup --add-users <file_name>\n```\nwhere the file\\_name consists of a list of user names and email addresses with the \nformat:\n```\nelmo elmo@gmail.com\nbigbird bigbird@gmail.com\n...\n```\n\nA message will be sent to the user's email address with their user name and a\ntemporary password, which will need to be changed when first attempting to\nsubmit a job.\n\nNote that the number of users able to be added each day is limited to 50 due to\nemail limits in the Cognito service. If you need to add more than this you\nwill need to configure Cognito with the Simple Email Service' (SES) and add a\nvalidated administrator email.\n\nThe cluster administrator will need to provide the server details to each \nuser which can be obtained by running the command:\n```\nqcloud_setup --user-info\nAwsRegion us-east-1\nCognitoUserPoolId us-east-1_KbkdtpKpW\nCognitoAppClientId 2mgcn0o8fkakboq7jnqs6bd6ee\nApiGatewayId fkkxolpuo4\n```\nIf these values are saved to a file:\n```\nqcloud_setup --user-info > user_info.txt\n```\nthen the cluster users can use this file to configure the client:\n```\nqcloud --configure user_info.txt\n```\nCluster users will need to install and configure the command line interface (CLI):\n```\npython3 -m pip install qcloud_user\n```\nthis will install the qcloud command into their python environment and this can be configured\nby running the following command and entering the values from the \n`qcloud_setup --user-info` command\n```\nqcloud --configure\n```\nThe first time a user interacts with the cluster they will need to enter the temporary\npassword emailed to them before resetting their password.\n\nSee the README.md file distributed with the qcloud\\_user module for further\ndetails on interacting with the server.\n\n\n\n\n## Suspending a Cluster\n\nIt is possible to shut down the cluster head node and restart it at a later\ntime in order to minimise the running costs. If you plan to restart the\ncluster, make sure to keep copies of the configuration file, ssh key (.pem\nfile) and activation key.\n\nUse the `--suspend` option to delete only the cluster stack. You will be prompted\nif you want to delete the stack, type 'y' to confirm.\n\n```\nqcloud_setup --suspend\nDelete stack qcloud-cluster? [y/N] y\n```\nThis terminates the head node, but leaves the API gateway and Cognito stacks\nrunning, using minimal resources. Once the cluster has been deleted, the\nresults of previous jobs can still be accessed via the API gateway as these are\narchived in an S3 bucket. You will, of course, be unable to submit further\njobs until the cluster has been restarted.\n\nTo re-launch the cluster, issue the following command in the same directory as\nthe configuration and license files:\n\n```\nqcloud_setup --launch \nqcloud_setup --activation-key XXXX-XXXX-XXXX-XXXX-XXXX\n```\n\n**Note:** In order to restart a cluster, you must have configured the cluster to\nuse an elastic IP (EIP) address allocated to your AWS account. This EIP must be\navailable for reuse when re-launching the cluster, otherwise the Q-Chem license\nwill no longer be valid.\n\nIf you need to update the EIP of your host, please contact our support team at \nsupport@q-chem.com for assistance.\n\n\n## Teminating a cluster\nBefore terminating a cluster, ensure there are no jobs running in the queue and\nthat you have downloaded the results of any calculations you wish to keep. The\nfollowing command will delete the cluster:\n```\nqcloud_setup --delete\n```\nYou will be asked whether you want to delete each of the 3 stacks, \nand you should type 'y' for each.\n\nTo clean up all the resources allocated to the cluster, be sure to release the\nelastic IP address and clean out any files created in the S3 bucket.\n\n\n## Other options\n\nAdditional options for the setup script can be printed via:\n\n```\nqcloud_setup --help\nqcloud_setup --info\nqcloud_setup --list\n```\n\nYou can also open an ssh connection to the head node using the following:\n```\nqcloud_setup --shell\n```\n\n## Costs\n\nThe costs associated with running the cluster hardware will be dominated by the\ncompute nodes used to run calculations. These costs depend on the type of node\nand the region configured during setup and can be estimated using the \n[AWS Cost Estimator](https://calculator.aws/#/addService/ec2-enhancement).\n\nIn addition to the compute-node costs, there are overhead costs associated with\nrunning the head-node used for job submissions which will be incurred even if\nthere are no running jobs. A low-cost T2 instance is used for this purpose and\nwill attract a monthly cost of around $10, depending on the amount of job\nstorage selected at setup time.\n\nThe costs associated with the head-node can be minimized by suspending the\ncluster when not in use. If this is done, there will still be a residual\ncharge for maintaining the elastic IP which is required for licensing. This\ncost is approximately $3.65 per month. \n\n*Note*: These costs are managed by AWS and will be charged to the account used\nto launch the Q-Cloud cluster, they are separate from the Q-Cloud subscription\nfee which is charged by Q-Chem Inc.\n\nTo monitor the costs associated with the cluster, the following will\ngive a summary of the costs recorded for the last 28 days:\n```\nqcloud_setup --costs\n```\nYou can also set notifications for when expenditure reaches 75% and 99% of\na specified budget, or when the expenditure is predicted to exceed the \nbudgeted figure:\n```\nqcloud_setup --set-budget n\n```\nwhere n is a integer number of USD. \n\n*Note*: Currently these cost settings apply to the entire AWS account.\n\n\n\n\n\n## Troubleshooting\n\n- When launching the cluster, if you receive permission problems associated\n with lambda functions, check to ensure that the IAM policy (QCloudIamPolicy) has been\n created in the same region as the cluster.\n\n- When submitting jobs, a job ID should be returned. This is a random\n sequence of 12 alphanumeric characters associated with your job. If you do\n not see a job ID, it is possible the submit function has not been updated\n with the instance ID of the head node. Re-run the launch command to trigger\n this update:\n```\nqcloud_setup --launch\n```\n\n- If you encounter any problems not covered here, please contact our support team at \n support@q-chem.com for assistance. \n\n",
"bugtrack_url": null,
"license": "Confidential property of Q-Chem, Inc. Copyright (c) 1993 by Q-Chem, Inc. (unpublished) All rights reserved. The above copyright notice is intended as a precaution against inadvertent publication and does not imply publication or any waiver of confidentiality. The year included in the foregoing notice is the year of creation. This software product contains proprietary, confidential information and trade secrets of Q-Chem, Inc. and its licensors. No use may be made of this software except according to written agreement with Q-Chem, Inc. ",
"summary": "Q-Cloud setup utility for cluster administrators",
"version": "1.0.33",
"project_urls": {
"Homepage": "https://q-chem.com"
},
"split_keywords": [
"qcloud",
" qchem",
" q-cloud",
" q-chem"
],
"urls": [
{
"comment_text": "",
"digests": {
"blake2b_256": "45979c3c33cc9edc7750c83ffe9e2d9c11319d2a38fdeaac313392c6fab1ef9f",
"md5": "6bada0218cd7d33cf857f029bed5bce4",
"sha256": "0d966571fa44c79106ab721feb1da8f361a30e7141681d098b9d8a15e4a9a256"
},
"downloads": -1,
"filename": "qcloud_setup-1.0.33-py3-none-any.whl",
"has_sig": false,
"md5_digest": "6bada0218cd7d33cf857f029bed5bce4",
"packagetype": "bdist_wheel",
"python_version": "py3",
"requires_python": ">=3.9",
"size": 52327,
"upload_time": "2024-04-19T00:25:17",
"upload_time_iso_8601": "2024-04-19T00:25:17.085113Z",
"url": "https://files.pythonhosted.org/packages/45/97/9c3c33cc9edc7750c83ffe9e2d9c11319d2a38fdeaac313392c6fab1ef9f/qcloud_setup-1.0.33-py3-none-any.whl",
"yanked": false,
"yanked_reason": null
},
{
"comment_text": "",
"digests": {
"blake2b_256": "c5dc088b1229469bf8c0defee84cb1515a1048d27379a89268e64b8ff244d8c2",
"md5": "781d69ecb4eb57867ff9850007e6e534",
"sha256": "52614dccc663c8148e0b79c76fbddc8fd23567edfaeca7e02123d0a97eb3faa7"
},
"downloads": -1,
"filename": "qcloud_setup-1.0.33.tar.gz",
"has_sig": false,
"md5_digest": "781d69ecb4eb57867ff9850007e6e534",
"packagetype": "sdist",
"python_version": "source",
"requires_python": ">=3.9",
"size": 34150,
"upload_time": "2024-04-19T00:25:19",
"upload_time_iso_8601": "2024-04-19T00:25:19.130885Z",
"url": "https://files.pythonhosted.org/packages/c5/dc/088b1229469bf8c0defee84cb1515a1048d27379a89268e64b8ff244d8c2/qcloud_setup-1.0.33.tar.gz",
"yanked": false,
"yanked_reason": null
}
],
"upload_time": "2024-04-19 00:25:19",
"github": false,
"gitlab": false,
"bitbucket": false,
"codeberg": false,
"lcname": "qcloud-setup"
}
JSON