CLI Tutorial: Container Volumes

Container (Docker) Volumes management functionality at the platform allows creating a persistent local storage, designed to be not affected by various operations during container lifecycle. Creation of such a storage right inside a container helps to ensure high level of integrity and prevent loss of important data.

Using platform CLI, you can set up container volumes in two ways:

• within topology settings
• via direct CLI methods

Set Volumes via Environment Topology

The first way to create container volume we’ll consider is to define it right during the environment creation (or when changing its topology). So, in order to setup your local container storage, the following parameters can be used while executing the appropriate CLI method:

• volumes - sets list of local volumes
• volumeMounts - allows to configure mount points
• volumesFrom - imports existing volumes from node on your account

1. All of these parameters are to be declared within the appropriate docker section. For example, the first parameter can be set in the following way:

1

... "docker":{... "volumes": ["{local_volume}", "{local_volume}", ...]}

Here, {local_volume} is a placeholder for the path your local volume should be placed inside a container. You can specify multiple volumes at once using comma separator.

E.g., within the image above, we create an environment with the jelastic/haproxy image inside and two local volumes attached, namely: /my_volume_1 and /my_volume_2.

2. The second volumeMounts parameter has the next syntaxis:

1

... "docker":{... "volumeMounts" : {"{local_path}" : {"sourcePath" : "{remote_path}", "sourceNodeId" : "{node_ID}", "readOnly" : {true/false}}, ...}}

The highlighted data should be changed to the appropriate values:

• {local_path} - path to the folder within nodes of the specified environment layer the mounted content should be placed at

• {remote_path} - the required directory location at the remote container (i.e. address of source data)

• {node_ID} - unique identifier of the node your data will be mounted from

  Note: Alternatively to the “sourceNodeId” parameter, you can also use:

  • “sourceNodeGroup” - defines environment layer, which master node will be used as a source for mount operation
  • “sourceHost” - external IP or custom domain name of the data storage server (including external one)

• {true/false} - the appropriate values define read only or read & write rights for client nodes while operating mounted data (false by default)

As an example, using the command on image above, we’ve mounted data from the 111109 node of the used account.

3. With the last volumesFrom parameter, it’s possible to mount all the existing volumes from one node to another (i.e. copy them), e.g.:

1

... "docker":{... "volumesFrom":[{"sourceNodeId" : "{node_ID}", "readOnly" : {true/false}, "volumes" : ["{local_volume}", "{local_volume}", ...]}]}

Here:

• {node_ID} - unique identifier of the node the data should be mounted from

  Note: Alternatively, it is possible to use “sourceNodeGroup” to define the environment layer, which master node will be used as a source for mount operation.

• {true/false} - the appropriate values define read only or read & write rights for client nodes while operating mounted data (false by default)

• {local_volume} - list of volumes to be mounted (if not specified all of them will be copied)

In such a way, using the command in the image above, we’ve added a new layer with two jelastic/tomcat8 Docker images and mounted the volumes to the existing jelastic/haproxy node.

Direct Volumes Management

The platform provides you with a number of CLI methods, that are developed specially for volumes management and do not affect the rest of environment topology:

1. The first one is the AddContainerVolume command, which lets you add a new volume to the existing environment:

1

~/jelastic/environment/control/addcontainervolume --envName {env_name} --nodeId {node_ID} --path {path}

where:

• {env_name} - domain name of the environment you’d like to modify
• {node_ID} - unique identifier of the node in chosen environment, a new container volume should be added to
• {path} - path in local filesystem to your volume directory

2. The removal process is absolutely similar to the just described addition process and uses the same parameters:

1

~/jelastic/environment/control/removecontainervolume --envName {env_name} --nodeId {node_ID} --path {path}

3. In case you need to add volumes not to a single node, but to a whole environment layer at once, the AddContainerVolumeByGroup method will be more suitable:

1

~/jelastic/environment/control/addcontainervolumebygroup --envName  {env_name} --nodeGroup  {node_group} --path  {path}

Here, the {node_group} parameter allows to define a layer in the current environment (e.g. cp, bl, storage, etc.), which nodes should be modified with a new volume.

4. Surely, you can remove a volume for the whole layer in a single command as well:

1

~/jelastic/environment/control/removecontainervolumebygroup --envName  {env_name} --nodeGroup  {node_group} --path  {path}

In such a way, using these CLI methods, you can easily take control over your container volumes and manage them in a matter of minutes. And in case you are in need of adding some Mount Points alongside with volumes, use the linked guide for detailed instruction.

What’s next?

Follow the linked guides for more platform CLI tutorials:

• environment creation
• environment start/stop
• environment cloning
• environment migration
• server scaling
• container redeploy
• mount points
• VCS projects deployment
• swap Public IPs
• installing JPS