Copyright (c) 2013-2015 CloudRunner.IO
CloudRunner.IO is a simple yet powerful framework for remote server management. It's key features include:
Execute scripts in your choice(bash, python, ruby, puppet and chef-solo recipes)
Pass environment variables between servers and between different script languages (works for a limited number of languages, but can be extended with a plugin)
Secure communication using SSL certificates
Fast and reliable, using ZeroMQ as a transport backend(although a different backend can be used as a plugin)
CloudRunner.IO also offers a commercial server, that performs a lot of useful tasks, including:
- User management: assign roles for different users on different remote servers(how to impersonate an user on a server)
- Library management: store, re-use scripts into different kind of stores - GitHub, ButBucket, SVN, Dropbox, Google Drive, Microsoft OneDrive, etc.
- Workflow management - run multi-step scripts on different servers, with the ability to restart a script from arbitrary step, while keeping the environment context as it was in the first run.
- Web dashboard for performing different operational tasks and for monitoring latest activities using filters.
- Execution of scheduled tasks (using Cron)
- Multi-tenancy - supports isolated group of users who can access servers in a shared environment (including public clouds).
- HA and Multi-server routing - install master servers in different locations(subnets, public clouds, etc.) and access all your servers from a single access point. No need to attach to different master server to access a remote server into directly inaccessible network. All you need is to allow the master servers to see each other.
- Highly customizable platform - write your own plugins(in Python) for different kind of workflow management.
For more details see www.cloudrunner.io or ask for details at [email protected]
CloudRunner Agent is an open-source project under the Apache 2 license. See the code at www.github.com/cloudrunner. Everyone is welcome to contribute.
Install and configure the Agent for use with CloudRunner.IO Master server:
pip install cloudrunner cloudrunner-node configure --org=MY-API-KEY -i NODE_NAME
Note: if NODE_NAME is skipped - the machine hostname will be used instead.
Note: you might need to install some packages before installing with pip. Cloudrunner depends on ZeroMQ, M2Crypto and httplib2. Install them using:
pip install pyzmq pip install m2crypto pip install httplib2
Note: Use python2 and pip2(python2-pip) for Arch Linux, CloudRunner only supports Python 2 at the moment, but this will change in future!
You can install them usign pip, but make sure you have already installed:
* C++ compiler: gcc-c++ (CentOS, Fedora) or gcc (Arch Linux) or g++ (Debian, Ubuntu) * Python Dev libraries: python-devel (CentOS, Fedora) or python2 (Arch Linux) or python-dev (Debian) * OpenSSL Dev libraries: libssl-dev (Debian) openssl-devel (Centos) or openssl (Arch Linux) * Swig package (swig) on some Linuxes
To use the Python CloudRunner.IO REST API client - install cloudrunner using pip:
pip install cloudrunner
Instantiate the client object:
from cloudrunner.api.client import *
client = Client('myusername', 'my_api_token')
Now you are ready to load/modify data on the server. Lets start with just listing the repositories in the Library:
repos = client.library.repositories.list() print repos
Returns an array of repositories:
[<cloudrunner.api.library.Repository at 0x7fb4f5e75fd0>, <cloudrunner.api.library.Repository at 0x7fb4f6dd7410>, <cloudrunner.api.library.Repository at 0x7fb4f5e777d0>, <cloudrunner.api.library.Repository at 0x7fb4f5e77890>]
You can use the returned objects and get their properties:
print repos[0].name >> "RepoName"
Lets see the contents of the root folder in first repo:
print client.library.browser.list(repos[0].name, '/') [<cloudrunner.api.library.Folder at 0x7f046c81cd90>, <cloudrunner.api.library.Script at 0x7f046c81ce10>, <cloudrunner.api.library.Script at 0x7f046c81cb90>, <cloudrunner.api.library.Script at 0x7f046c81cc10>, <cloudrunner.api.library.Script at 0x7f046c81cc50>, <cloudrunner.api.library.Script at 0x7f046c81cc90>, <cloudrunner.api.library.Script at 0x7f046c81cbd0>]
Now lets make a search for a specific log:
logs = client.logs.search.list(filter='my favourite pattern') print logs
Fortunatelly, we get some results:
[<cloudrunner.api.logs.Log at 0x7f046c81cfd0>, <cloudrunner.api.logs.Log at 0x7f046c81ca10>, <cloudrunner.api.logs.Log at 0x7f046c81cb10>, <cloudrunner.api.logs.Log at 0x7f046c81cb50>, <cloudrunner.api.logs.Log at 0x7f046c81f090>, <cloudrunner.api.logs.Log at 0x7f046c81f450>]
To retrieve the information for a log, we will load it using:
log = client.logs.get.item(logs[0]) print log >> <cloudrunner.api.logs.Log object at 0x7f046c7b5310>
To get the runs under a specific workflow in a Log:
print log.workflow[0].runs[0] << [<cloudrunner.api.base.ApiObject at 0x7f046c7b5d90>] print log.workflow[0].runs[0].uuid << u'ff57f0b8ac1a426783d5763626be07cb'
We want to see the output from the first run in the first workflow:
logs = client.logs.output.item(l.workflows[0].runs[0].uuid)
print logs
>> [<cloudrunner.api.logs.Log at 0x7f046d08f390>]
print logs[0].screen
>> <cloudrunner.api.base.ApiObject object at 0x7f046c7b5f90>
print logs[0].screen._values
>> {u'NODE_NAME': {u'lines': [[1423063412.399, [u'OUTPUT FROM MY SCRIPT EXECUTION'], u'O']]}}
In fact, we can load the node data directly from the screen object:
print logs[0].screen.NODE_NAME.lines >> [[1423063412.399, [u'OUTPUT FROM MY SCRIPT EXECUTION'], u'O']]
We can also get the currently registered nodes in our account:
nodes = client.nodes.nodes.list()
print nodes
>> [<cloudrunner.api.nodes.Node at 0x7f8664b10250>,
<cloudrunner.api.nodes.Node at 0x7f86645151d0>,
<cloudrunner.api.nodes.Node at 0x7f8664515050>]
node = nodes[0]
print node.name
>> NODE_NAME
print node.meta._values
>> {
u'ARCH': u'x86_64',
u'AVAIL_MEM': 767,
u'CPUS': None,
u'CPU_CORES': 1,
u'CRN_VER': u'1.1.0',
u'DIST': u'CentOS',
u'HOST': u'NODE_NAME',
u'ID': u'NODE_NAME',
u'MASTER_IP': u'192.168.1.1',
u'OS': u'Linux',
u'PRIVATE_IP': [u'10.1.1.1', u'127.0.0.1'],
u'PUBLIC_IP': [u'54.1.1.1'],
u'RELEASE': u'2.6.32-431.29.2.el6.x86_64',
u'SERVER_NAME': u'NODE_NAME',
u'TOTAL_MEM': 992
}
print node.joined_at
>> u'2015-01-28 10:39:44'