Migrating a Deployment to a New Set of Hosts

This section describes how to migrate a Cloudera Data Science Workbench deployment to a new set of gateway hosts.

Migrating a CSD Deployment

This section describes how to migrate a CSD-based Cloudera Data Science Workbench service to a new set of gateway hosts.

  1. Add and Set Up the New Hosts
  2. Stop the CDSW Service
  3. Backup Application Data
  4. Delete CDSW Roles from Existing Hosts
  5. Move Backup to the New Master
  6. Update DNS Records for the New Master
  7. Add Role Instances for the New Hosts
  8. Run the Prepare Node command on the New Hosts
  9. Start the CDSW Service

Add and Set Up the New Hosts

  1. Add new hosts to your cluster as needed. Make sure they are gateway hosts that have been assigned gateway roles for HDFS, YARN, and Spark 2. Do not run any other services on these hosts.

  2. Set up the new hosts as per the Cloudera Data Science Workbench hardware requirements listed here.

Stop the CDSW Service

Use Cloudera Manager to stop all roles of the CDSW service.
  1. Log into the Cloudera Manager Admin Console.
  2. On the Home > Status tab, click to the right of the CDSW service and select Stop from the dropdown.
  3. Confirm your choice on the next screen. When you see a Finished status, the action is complete.

Backup Application Data

In Cloudera Data Science Workbench all stateful data is stored on the master host at /var/lib/cdsw. Backup the contents of this directory before you begin the migration process.

To create the backup, run the following command on the master host: 
tar -cvzf cdsw.tar.gz -C /var/lib/cdsw/ .

Delete CDSW Roles from Existing Hosts

  1. Log into the Cloudera Manager Admin Console.
  2. Go to the CDSW service.
  3. Click the Instances tab.
  4. Select all the role instances.
  5. Select Actions for Selected > Delete. Click Delete to confirm the deletion.

Move Backup to the New Master

Copy the backup taken previously to the host that will be the new Cloudera Data Science Workbench master. Unpack the contents of the backup into /var/lib/cdsw. 
tar xvzf cdsw.tar.gz -C /var/lib/cdsw

Update DNS Records for the New Master

Update your DNS records with the IP address for the new master host.

Add Role Instances for the New Hosts

  1. Log into the Cloudera Manager Admin Console.
  2. Go to the CDSW service.
  3. Click the Instances tab.
  4. Click Add Role Instances. Assign the Cloudera Data Science Workbench Master, Application, and Docker Daemon roles to the new master host. If you want to configure worker hosts, assign the Cloudera Data Science Workbench Worker and Docker Daemon roles to the new workers.
  5. Click Continue. On the Review Changes page, review the configuration changes to be applied. The wizard finishes by performing any actions necessary to add the new role instances.

    Do not start the new roles at this point. You must run the Prepare Node command as described in the next step before the roles are started.

Run the Prepare Node command on the New Hosts

The new hosts must have the following packages installed on it. 
nfs-utils
libseccomp
lvm2
bridge-utils
libtool-ltdl
iptables   
rsync 
policycoreutils-python 
selinux-policy-base 
selinux-policy-targeted 
ntp 
ebtables 
bind-utils 
nmap-ncat  
openssl 
e2fsprogs 
redhat-lsb-core 
socat
You can either manually install these packages now, or, allow Cloudera Manager to install them as part of the Prepare Node command later in this step.

If you choose the latter, make sure that Cloudera Manager has the permissions needed to install the required packages. To do so, go to the CDSW service and click Configuration. Search for the Install Required Packages property and make sure it is enabled.

Then run the Prepare Node command on the new hosts.
  1. Go to the CDSW service.
  2. Click Instances.
  3. Select all the role instances.
  4. Select Actions for Selected > Prepare Node. This will install the required set of packages on all the new hosts.

Start the CDSW Service

  1. Log into the Cloudera Manager Admin Console.
  2. On the Home > Status tab, click to the right of the CDSW service and select Start from the dropdown.
  3. Confirm your choice on the next screen. When you see a Finished status, the action is complete.

Migrating an RPM Deployment

This section describes how to migrate an RPM-based Cloudera Data Science Workbench service to a new set of gateway hosts.

  1. Add and Set Up the New Hosts
  2. Stop Cloudera Data Science Workbench
  3. Backup Application Data
  4. Remove Cloudera Data Science Workbench from Existing Hosts
  5. Move Backup to New Master
  6. Update DNS Records for the New Master
  7. Install Cloudera Data Science Workbench on New Master Host

Add and Set Up the New Hosts

  1. Add new hosts to your cluster as needed. Make sure they are gateway hosts that have been assigned gateway roles for HDFS, YARN, and Spark 2. Do not run any other services on these hosts.

  2. Set up the new hosts as per the Cloudera Data Science Workbench hardware requirements listed here.

Stop Cloudera Data Science Workbench

Run the following command on the master host to stop Cloudera Data Science Workbench. 
cdsw stop

Backup Application Data

In Cloudera Data Science Workbench all stateful data is stored on the master host at /var/lib/cdsw. Backup the contents of this directory before you begin the migration process.

To create the backup, run the following command on the master host: 
tar -cvzf cdsw.tar.gz -C /var/lib/cdsw/ .

Remove Cloudera Data Science Workbench from Existing Hosts

Run the following commands on the existing master and any worker hosts you want to migrate. 
cdsw reset
yum remove cloudera-data-science-workbench

Move Backup to New Master

Copy the backup taken in the previous step to the host that will be the new Cloudera Data Science Workbench master. Unpack the contents of the backup into /var/lib/cdsw. 
tar xvzf cdsw.tar.gz -C /var/lib/cdsw

Update DNS Records for the New Master

Update your DNS records with the IP address for the new master host.

Install Cloudera Data Science Workbench on New Master Host

For instructions, see Installing Cloudera Data Science Workbench 1.5.x Using Packages.