Coriolis CLI - Cloudbase Solutions

This section will provide an overview on how to use the CLI to interact with Coriolis.

For an overview of the functioning and mechanics of Coriolis, please review the Product Overview page here.

Due to Coriolis’ pluggable nature, this guide presents general interactions with Coriolis, and apply to work with any source/destination cloud Coriolis supports. Please review the documentation of each respective cloud’s Coriolis plugin to get exact details on the required fields of parameters such as the – -connection or – -destination-environment

The below outlines running the Coriolis CLI client from the Coriolis appliance. Coriolis CLI client can be installed and executed from other systems that have access to Coriolis.

Installing the Coriolis command line client

The following commands will install the Coriolis command line client on the local machine.

Coriolis CLI installation commands

$ git clone https://github.com/cloudbase/python-coriolisclient

1	$ git clone https://github.com/cloudbase/python-coriolisclient

$ pip3 install python-coriolisclient
$ # verify client is available in path:
$ /usr/bin/env coriolis

$ pip3 install python-coriolisclient

$ # verify client is available in path:

$ /usr/bin/env coriolis

Expose Coriolis appliance

By default, all the Coriolis appliance services listen only on localhost. In order to expose all the services from the appliance, execute the following script on the Coriolis Appliance:

$ ~/coriolis-docker/expose_coriolis.py

1	$ ~/coriolis-docker/expose_coriolis.py

Apply RC file so you can run the Coriolis commands from where the Coriolis client has been installed

By default, the Coriolis RC file for the admin account is to be found under /etc/kolla/ on the Coriolis Appliance

#select option '3) Edit/Inspect Coriolis Configuration'

1	#select option '3) Edit/Inspect Coriolis Configuration'

$cat /etc/kolla/admin-openrc.sh

1	$cat /etc/kolla/admin-openrc.sh

It will also need to be copied on the machine where the Coriolis client has been installed.

$scp /etc/kolla/admin-openc.sh adrian@<IP address>:/mnt/c/coriolis/

1	$scp /etc/kolla/admin-openc.sh adrian@<IP address>:/mnt/c/coriolis/

After copying it, the file needs to be sourced on the machine where the client runs and Coriolis commands can be performed afterwards

$source admin-openrc.sh

$coriolis endpoint list

$source admin-openrc.sh

$coriolis endpoint list

Defining a Coriolis endpoint for a given cloud

In order to perform the steps needed to migrate/replicate to/from a given platform, Coriolis will need credentials for accessing the said platform.

Coriolis endpoints are a type of resource definable in Coriolis which simply declares the type and connection parameters for a given cloud platform. The connection parameters must be in JSON format, the exact structure of which depends on the type of cloud Coriolis is interacting with. Please review the documentation of the plugin for the respective cloud you are dealing with to see the format needed for the connection info JSON.

Below is an example for creating an endpoint for a fictive “CloudX”

Endpoint creation example

#the example of the RC file is reffering to the one on the Coriolis Appliance
$source /etc/kolla/admin-openrc.sh
$coriolis endpoint create \
      --name cloudx --description "CloudX endpoint" \
      --provider cloudx --connection '{"username": "iamtheowner", "password": "SeKr37"}'

#the example of the RC file is reffering to the one on the Coriolis Appliance

$source /etc/kolla/admin-openrc.sh

$coriolis endpoint create \

--name cloudx --description "CloudX endpoint" \

--provider cloudx --connection '{"username": "iamtheowner", "password": "SeKr37"}'

The cloud plugin identifier given through the –provider parameter is unique to each plugin. Please review the documentation of the respective platform’s Coriolis plugin to find the correct plugin identifier for it.

Later on when we will want to start migration/replica jobs to/from CloudX, we will be able to simply reference the endpoint created above by its ID instead of having to fill in the credentials of the clouds involved at each and every new task.

Storing Coriolis endpoint connection info in Barbican

Coriolis features integration with Barbican, the OpenStack project dedicated to the storage and management of sensitive information.

Instead of providing the plaintext JSON –connection directly to Coriolis, one might first store it in Barbican and only provide Coriolis with a reference to it.

Endpoint creation example with Barbican secret

#the example of the RC file is reffering to the one on the Coriolis Appliance
$source /etc/kolla/admin-openrc.sh
$openstack secret store --name "CloudX credentials" -t "text/plain" \
 -p '{"username": "iamtheowner", "password": "SeKr37"}'
$coriolis endpoint create \
 --name cloudx --description "CloudX endpoint" \
 --provider cloudx --connection-secret "$URL_OF_ABOVE_BARBICAN_SECRET"

#the example of the RC file is reffering to the one on the Coriolis Appliance

$source /etc/kolla/admin-openrc.sh

$openstack secret store --name "CloudX credentials" -t "text/plain" \

-p '{"username": "iamtheowner", "password": "SeKr37"}'

$coriolis endpoint create \

--name cloudx --description "CloudX endpoint" \

--provider cloudx --connection-secret "$URL_OF_ABOVE_BARBICAN_SECRET"

Barbican will encrypt the connection info before storing it into the database and Coriolis will only access the credentials and pass them to the respective Coriolis platform plugin when it uses its services.

Validating a Coriolis endpoint

Once a cloud endpoint has been created for the specific platform we are dealing with, Coriolis may be instructed to validate that the provided credentials are correct by performing a test with the credentials.

From the command line client, this may be done via running the following:

Endpoint validation example

#the example of the RC file is reffering to the one on the Coriolis Appliance
$source /etc/kolla/admin-openrc.sh
$coriolis endpoint validate connection $ENDPOINT_ID

#the example of the RC file is reffering to the one on the Coriolis Appliance

$source /etc/kolla/admin-openrc.sh

$coriolis endpoint validate connection $ENDPOINT_ID

Should the credentials be in order, the command will terminate with no output and error code 0. Otherwise, a descriptive error message received back from the Coriolis service will be displayed.

Listing instances available for migration/replication from a source platform

If Coriolis supports migrating/replicating from a particular platform, then the Coriolis plugin for that platform implements functionality for listing the instances available for the operation.

From the command line client, the following may be run to list instances available for export:

Endpoint instance list example

#the example of the RC file is reffering to the one on the Coriolis Appliance
$source /etc/kolla/admin-openrc.sh
$coriolis endpoint instance list --source-environment '{"source": "parameters"}' $ENDPOINT_ID

#the example of the RC file is reffering to the one on the Coriolis Appliance

$source /etc/kolla/admin-openrc.sh

$coriolis endpoint instance list --source-environment '{"source": "parameters"}' $ENDPOINT_ID

The parameters for the –source-environment are source cloud specific and will aid Coriolis in listing instances (e.g: ‘resource_group’ for Azure/AzureStack). Please review the documentation of the Coriolis platform plugin of the source cloud for the exact parameters which are expected.

For starting migrations/replicas for a given instance, whether or not it may be identified via its name and/or its ID depends on the implementation of the Coriolis platform plugin for that particular cloud. Please review the documentation of the respective platform’s Coriolis plugin for exact instance referencing requirements.

Additional operations for Coriolis endpoints

In addition to the above notable commands, the Coriolis command line client also offers the following operations on cloud endpoints:

Additional endpoint operations examples

#the example of the RC file is reffering to the one on the Coriolis Appliance
$source /etc/kolla/admin-openrc.sh
 #list all defined endpoints:
$coriolis endpoint list
 #detailed view of a specific endpoint:
$coriolis endpoint show $ENDPOINT_ID
 #update the parameters of a specific endpoint:
$coriolis endpoint update --param "updated value for --param" $ENDPOINT_ID
 #delete a specific endpoint:
$coriolis endpoint delete $ENDPOINT_ID

#the example of the RC file is reffering to the one on the Coriolis Appliance

$source /etc/kolla/admin-openrc.sh

#list all defined endpoints:

$coriolis endpoint list

#detailed view of a specific endpoint:

$coriolis endpoint show $ENDPOINT_ID

#update the parameters of a specific endpoint:

$coriolis endpoint update --param "updated value for --param" $ENDPOINT_ID

#delete a specific endpoint:

$coriolis endpoint delete $ENDPOINT_ID

Creating migration jobs

Prerequisites:

an existing Coriolis endpoint for the migration source cloud (granted the source cloud has a Coriolis platform plugin available)
an existing Coriolis endpoint for the migration destination cloud (granted the source cloud has a Coriolis platform plugin available)
the parameters for the –source-environment, which are source cloud specific. Please review the documentation of the Coriolis platform plugin of the source cloud for the exact parameters which are expected
the parameters for the –destination-environment, which are destination cloud specific. Please review the documentation of the Coriolis platform plugin of the destination cloud for the exact parameters which are expected
a unique identifier of the instance we would like to migrate

Granted all the above, a migration job may be created for an instance by running the following:

Creating a Coriolis migration

#the example of the RC file is reffering to the one on the Coriolis Appliance
$source /etc/kolla/admin-openrc.sh
$coriolis migration create \
 --origin-endpoint $ID_OF_SOURCE_CLOUD_ENDPOINT \
 --destination-endpoint $ID_OF_DESTINATION_CLOUD_ENDPOINT \
 --source-environment '{"the": "source cloud specific params"}' \
 --destination-environment '{"the": "destination cloud specific params"}' \
 --network-map '{"source network name": "name or ID of network on destination"}' \
 --instance $IDENTIFIER_OF_INSTANCE_ON_SOURCE

#the example of the RC file is reffering to the one on the Coriolis Appliance

$source /etc/kolla/admin-openrc.sh

$coriolis migration create \

--origin-endpoint $ID_OF_SOURCE_CLOUD_ENDPOINT \

--destination-endpoint $ID_OF_DESTINATION_CLOUD_ENDPOINT \

--source-environment '{"the": "source cloud specific params"}' \

--destination-environment '{"the": "destination cloud specific params"}' \

--network-map '{"source network name": "name or ID of network on destination"}' \

--instance $IDENTIFIER_OF_INSTANCE_ON_SOURCE

The above command will create a migration job within Coriolis, which may be later queried or halted by referencing its ID.

Additional operations for Coriolis migrations

In addition to the above create command, the Coriolis command line client also offers the following operations on migration jobs:

Additional endpoint operations examples

#the example of the RC file is reffering to the one on the Coriolis Appliance
$source /etc/kolla/admin-openrc.sh
 #list all registered migration jobs:
$coriolis migration list
 #view a particular migration's status:
$coriolis migration show $MIGRATION_ID
 #cancel a running migration:
$coriolis migration cancel $MIGRATION_ID
 #delete records of a previous migration:
 #this does not delete the migrated instance on the destination cloud
 #should that migration have been ended successfully.
 #the migration must not be running.
$coriolis migration delete $MIGRATION_ID

#the example of the RC file is reffering to the one on the Coriolis Appliance

$source /etc/kolla/admin-openrc.sh

#list all registered migration jobs:

$coriolis migration list

#view a particular migration's status:

$coriolis migration show $MIGRATION_ID

#cancel a running migration:

$coriolis migration cancel $MIGRATION_ID

#delete records of a previous migration:

#this does not delete the migrated instance on the destination cloud

#should that migration have been ended successfully.

#the migration must not be running.

$coriolis migration delete $MIGRATION_ID

Creating replication (DRaaS) jobs

Prerequisites:

an existing Coriolis endpoint for the replication source cloud (granted the source cloud has a Coriolis platform plugin available)
an existing Coriolis endpoint for the replication destination cloud (granted the source cloud has a Coriolis platform plugin available)
the parameters for the –source-environment, which are source cloud specific. Please review the documentation of the Coriolis platform plugin of the source cloud for the exact parameters which are expected
the parameters for the –destination-environment, which are destination cloud specific. Please review the documentation of the Coriolis platform plugin of the destination cloud for the exact parameters which are expected
a unique identifier of the instance we would like to replicate

Granted all the above, a replication job may be defined (but not yet executed) for an instance by running the following:

Creating a Coriolis replica

#the example of the RC file is reffering to the one on the Coriolis Appliance
$source /etc/kolla/admin-openrc.sh
$coriolis replica create \
 --origin-endpoint $ID_OF_SOURCE_CLOUD_ENDPOINT \
 --destination-endpoint $ID_OF_DESTINATION_CLOUD_ENDPOINT \
 --source-environment '{"the": "source cloud specific params"}' \
 --destination-environment '{"the": "destination cloud specific params"}' \
 --network-map '{"source network name": "name or ID of network on destination"}' \
 --instance $IDENTIFIER_OF_INSTANCE_ON_SOURCE

#the example of the RC file is reffering to the one on the Coriolis Appliance

$source /etc/kolla/admin-openrc.sh

$coriolis replica create \

--origin-endpoint $ID_OF_SOURCE_CLOUD_ENDPOINT \

--destination-endpoint $ID_OF_DESTINATION_CLOUD_ENDPOINT \

--source-environment '{"the": "source cloud specific params"}' \

--destination-environment '{"the": "destination cloud specific params"}' \

--network-map '{"source network name": "name or ID of network on destination"}' \

--instance $IDENTIFIER_OF_INSTANCE_ON_SOURCE

The above command will define a replication job with Coriolis, which may later have executions (syncs) scheduled, canceled, and deployed by referencing the replication job’s ID.

Starting an execution for a replication job

Prerequisites: an already defined replication job that does not already have a running execution

An existing replica may have a new execution (sync run) started by running the following:

Executing a Coriolis replica

#the example of the RC file is reffering to the one on the Coriolis Appliance
$source /etc/kolla/admin-openrc.sh
$coriolis replica execute $REPLICA_ID

#the example of the RC file is reffering to the one on the Coriolis Appliance

$source /etc/kolla/admin-openrc.sh

$coriolis replica execute $REPLICA_ID

The above command will launch a new replica execution, which may later be queried for status, canceled, or deleted by referencing both the ID of the Replica, and the ID of the newly-created replica execution.

Additional operations for Coriolis replica executions

In addition to the above execution command, the Coriolis command line client also offers the following operations on replica executions:

Additional replica execution operations examples

#the example of the RC file is reffering to the one on the Coriolis Appliance
$source /etc/kolla/admin-openrc.sh
 #list all executions of a replica:
$coriolis replica execution list $REPLICA_ID
 #view a particular replica execution's status:
$coriolis replica execution show $REPLICA_ID $REPLICA_EXECUTION_ID
 #cancel a running replica execution:
$coriolis replica execution cancel $REPLICA_ID $REPLICA_EXECUTION_ID
 #delete records of a non-running replica execution:
 #this will not delete any resources which might have been
 #replicated on the destination cloud if the replica execution
 #had completed successfully (such as the replica disks)
$coriolis replica execution delete $REPLICA_ID $REPLICA_EXECUTION_ID
 #delete the replica disks:
 #this will delete the disks on the destination cloud we are
 #replicating to, but it will not delete replica in Coriolis.
 #Later executions may be launched for the replica, but the next one
 #will be a full execution to completely new disks on the destination.
$coriolis replica disks delete $REPLICA_ID

#the example of the RC file is reffering to the one on the Coriolis Appliance

$source /etc/kolla/admin-openrc.sh

#list all executions of a replica:

$coriolis replica execution list $REPLICA_ID

#view a particular replica execution's status:

$coriolis replica execution show $REPLICA_ID $REPLICA_EXECUTION_ID

#cancel a running replica execution:

$coriolis replica execution cancel $REPLICA_ID $REPLICA_EXECUTION_ID

#delete records of a non-running replica execution:

#this will not delete any resources which might have been

#replicated on the destination cloud if the replica execution

#had completed successfully (such as the replica disks)

$coriolis replica execution delete $REPLICA_ID $REPLICA_EXECUTION_ID

#delete the replica disks:

#this will delete the disks on the destination cloud we are

#replicating to, but it will not delete replica in Coriolis.

#Later executions may be launched for the replica, but the next one

#will be a full execution to completely new disks on the destination.

$coriolis replica disks delete $REPLICA_ID

Deploying a replica on the destination cloud

Prerequisite: a pre-existing Coriolis replica that has at least one successful replica execution

An existing replica may be deployed on the destination cloud by running the following:

Deploying a Coriolis replica

#the example of the RC file is reffering to the one on the Coriolis Appliance
$source /etc/kolla/admin-openrc.sh
$coriolis migration deploy replica $REPLICA_ID

#the example of the RC file is reffering to the one on the Coriolis Appliance

$source /etc/kolla/admin-openrc.sh

$coriolis migration deploy replica $REPLICA_ID

The above command will start a replica migration job which will recreate the instance on the destination cloud with the state of the last replica execution.

Using the replica migration job’s ID to reference it, you may run the same “Coriolis migration …” operations detailed in the section on creating migration jobs.

Coriolis Replica/Migration task tutorial

For more information on how to create a Replica/Migration using Coriolis’ CLI, please check the Coriolis Replica/Migration task page.

Coriolis Minion Pools

Coriolis Minion Pools are a resource that relies on temporary virtual machines which it deploys on the source/target platforms to perform various actions during transfer operations during Coriolis Replica/Migration jobs.

For more information regarding Coriolis Minion Pools using CLI, please check Minion Pools – Creation and Management page.