API Reference

os_faults.connect(cloud_config=None, config_filename=None)

Connect to the cloud

Parameters:

• cloud_config – dict with cloud and power management params
• config_filename – name of the file where to read config from

Returns:

CloudManagement object

os_faults.discover(cloud_config)

Connect to the cloud and discover nodes and services

Parameters:cloud_config – dict with cloud and power management params Returns:config dict with discovered nodes/services

os_faults.human_api(distractor, command)

Execute high-level text command with specified destructor

Parameters:

• destructor – library instance as returned by :connect: function
• command – text command

os_faults.register_ansible_modules(paths)

Registers ansible modules by provided paths

Allows to use custom ansible modules in NodeCollection.run_task method

Parameters:path – list of paths to folders with ansible modules

class os_faults.api.cloud_management.CloudManagement

execute_on_cloud(hosts, task, raise_on_error=True)

Execute task on specified hosts within the cloud.

Parameters:

• hosts – List of host FQDNs
• task – Ansible task
• raise_on_error – throw exception in case of error

Returns:

Ansible execution result (list of records)

get_nodes(fqdns=None)

Get nodes in the cloud

This function returns NodesCollection representing all nodes in the cloud or only those that has specified FQDNs. :param fqdns list of FQDNs or None to retrieve all nodes :return: NodesCollection

get_service(name)

Get service with specified name

Parameters:name – name of the service Returns:Service

classmethod list_supported_networks()

Lists all networks supported by nodes returned by this driver

Returns:[String] list of network names

classmethod list_supported_services()

Lists all services supported by this driver

Returns:[String] list of service names

verify()

Verify connection to the cloud.

class os_faults.api.service.Service(service_name, config, node_cls, cloud_management, hosts=None)

discover_nodes()

Discover nodes where this Service is running

Returns:NodesCollection

freeze(nodes=None, sec=None)

Pause service execution

Send SIGSTOP to Service into network on all nodes or on particular subset. If sec is defined - it mean Service will be stopped for a wile.

Parameters:

• nodes – NodesCollection
• sec – int

get_nodes()

Get nodes where this Service is running

Returns:NodesCollection

kill(nodes=None)

Terminate Service abruptly on all nodes or on particular subset

Parameters:nodes – NodesCollection

plug(nodes=None)

Plug Service into network on all nodes or on particular subset

Parameters:nodes – NodesCollection

restart(nodes=None)

Restart Service on all nodes or on particular subset

Parameters:nodes – NodesCollection

start(nodes=None)

Start Service on all nodes or on particular subset

Parameters:nodes – NodesCollection

terminate(nodes=None)

Terminate Service gracefully on all nodes or on particular subset

Parameters:nodes – NodesCollection

unfreeze(nodes=None)

Resume service execution

Send SIGCONT to Service into network on all nodes or on particular subset.

Parameters:nodes – NodesCollection

unplug(nodes=None)

Unplug Service out of network on all nodes or on particular subset

Parameters:nodes – NodesCollection

class os_faults.api.node_collection.NodeCollection(cloud_management=None, hosts=None)

connect(network_name)

Connect nodes to <network_name> network

Parameters:network_name – name of network

disconnect(network_name)

Disconnect nodes from <network_name> network

Parameters:network_name – name of network

oom()

Fill all node’s RAM

pick(count=1)

Pick one Node out of collection

Returns:NodeCollection consisting just one node

poweroff()

Power off all nodes abruptly

poweron()

Power on all nodes abruptly

reboot()

Reboot all nodes gracefully

reset()

Reset (cold restart) all nodes

revert(snapshot_name, resume=True)

Revert snapshot for all nodes

run_task(task, raise_on_error=True)

Run ansible task on node colection

Parameters:

• task – ansible task as dict
• raise_on_error – throw exception in case of error

Returns:

AnsibleExecutionRecord with results of task

shutdown()

Shutdown all nodes gracefully

snapshot(snapshot_name, suspend=True)

Create snapshot for all nodes

stress(target, duration=None)

Stress node OS and hardware