Cluster Management Tool (CMT) is originally created at SARA Computing and Networking Services, which is based in Amsterdam and now known as SURFsara.
The main reason behind CMT's existence is that we needed a tool that's capable of generating configuration-files for certain software running on our clusters.
CMT uses a database to store all information about your networks, hardware and interfaces. With templates you can generate different types of configuration files.
For example DHCP, DNS, Torque, etc. CMT uses the Django framework.
The documentation is currently a bit limited to this README for know. We are working on better docmentation
- Python 2.7 (working on Python 3 support)
- Python virtualenv
All dependencies for CMT can be found in the requirements.txt file. If you don't want to use the LDAP authentication backend, then you can remove django-auth-ldap from the requirements.txt file.
Please note that CMT has been testen on Linux only.
After creating a virtualenv you can install the requirements of CMT with the following command:
pip install -r requirements.txt
Then you will need to create a configuration file which is used by CMT. CMT will look
at location /etc/cmt/cmt.conf. For an example configuration see files/cmt.conf .
For easy use you can create a file .virtual_env_path in the root dir of CMT django site. In this
file you will need to specify the full path to your virtualenv you have created for CMT. For example
when you created a virtualenv called cmt-2.5.0 the path is /home/<username>/.virtualenv/cmt-2.5.0.
Finally you will need to bootstrap CMT with ./manage.py migrate and finally you can start CMT with
./manage.py runserver.
Before you can use the client, you will need to download it first. You can do this from the web
interface or with curl or wget. The client has only 1 requirement, Python 2.7. We are working
Python 3 support. In the next release 2.6.0 Python 3 wil be fully supported.
Examples:
curl --user <username> https://<hostname>/client/download
wget --user=<username> --ask-password --auth-no-challenge https://<hostname>/client/download
The client does not need any configuration, it will be configured for you when you download the client. The client supports the following modes:
- read, Fetch and object from CMT
- create, Create an object in CMT
- update, Update an object from CMT
- parse, Parse a template from CMT
You can read, create, update, parse the following objects:
- address
- cluster
- company
- connection
- country
- equipment
- hardwaremodel
- interface
- interfacetype
- network
- rack
- role
- room
- telephonenumber
- warrantycontract
- warrantytype
The client uses the arguments --get and --set to determine what kind of fields are specified.
For each mode you can use the --help which of the arguments are supported by which mode.
some basics:
./cmt_client.py --help
sage: cmt_client.py [-h] [-v] [-n] [--version]
{read,create,update,delete,parse} ...
This is the api client tool for the cmt server
positional arguments:
{read,create,update,delete,parse}
read Fetch and object from CMT
create Create an object in CMT
update Update an object from CMT
delete Delete an object from CMT
parse Parse a template from CMT
optional arguments:
-h, --help show this help message and exit
-v, --verbose Enable verbose mode
-n, --dry-run Enable dry-run, also enables verbose mode
--version show program's version number and exit
subcommand help:
root# ./cmt_client.py parse --help
usage: cmt_client.py parse [-h] [-w] [-d] [-e] [filename]
positional arguments:
filename Specify the templatefile, leave empty for menu
optional arguments:
-h, --help show this help message and exit
-w, --write Do not ask to write, just write
-d, --skip-diff Do not ask to show the diff
-e, --skip-epilogue Skip epilogue
Create example:
python cmt_client.py create cluster --set name="CMT Cluster"
Username: <username>
Password:
[CREATE] Result:
{
"created_on": "2017-06-17T19:57:12.875774",
"machinenames": null,
"name": "CMT Cluster",
"note": "",
"response": "201 - CREATED",
"tags": "",
"updated_on": "2017-06-17T19:57:12.875822",
"url": "http://localhost.localdomain:8000/api/v2/cluster/55"
}
Read example:
python cmt_client.py read cluster --get name="CMT Cluster"
[READ] Result:
{
"count": 1,
"next": null,
"previous": null,
"response": "200 - OK",
"results": [
{
"created_on": "2017-06-17T19:57:12.875774",
"machinenames": null,
"name": "CMT Cluster",
"note": "",
"tags": "",
"updated_on": "2017-06-17T19:57:12.875822",
"url": "http://localhost.localdomain:8000/api/v2/cluster/55"
}
]
}
Update example:
python cmt_client.py update cluster --get name="CMT Cluster" --set machinenames="{machinenames}"
Username: <username>
Password:
[UPDATE] Result:
{
"count": 1,
"next": null,
"previous": null,
"results": [
{
"created_on": "2017-06-17T19:57:12.875774",
"machinenames": "{machinenames}",
"name": "CMT Cluster",
"note": "",
"response": "200 - OK",
"tags": "",
"updated_on": "2017-06-17T19:59:54.621700",
"url": "http://localhost.localdomain:8000/api/v2/cluster/55"
}
]
}
Delete example:
python cmt_client.py delete cluster --get name="CMT Cluster"
Username: <username>
Password:
[DELETE] Result:
{
"count": 1,
"next": null,
"previous": null,
"results": [
{
"created_on": "2017-06-17T19:57:12.875774",
"machinenames": "{machinenames}",
"name": "CMT Cluster",
"note": "",
"tags": "",
"updated_on": "2017-06-17T19:59:54.621700",
"url": "http://localhost.localdomain:8000/api/v2/cluster/55"
}
]
}
Read example:
import cmt_client
client = cmt_client.Client()
json = client.request('read', 'cluster', ['name="CMT Cluster"'], None)
print json- Currently field roles is not marked as required in the API when creating a new Equipment, but it is required in the database model. So you will need to remember to specify it.
- Currently is is not possible to add multiple roles to an Equipment via the API.
- CMT client is not Python 3 ready
For bugs, requests or suggestions, please mail to cmt-dev@surfsara.nl
History and old documentation can be found here: (https://oss.trac.surfsara.nl/cmt)