Skip to content

Latest commit

 

History

History
181 lines (121 loc) · 6.09 KB

how-to-migrate.adoc

File metadata and controls

181 lines (121 loc) · 6.09 KB
Raw

Migrating blueprint from Ravello to IBM Cloud Storage

As we know we are near close to EOL of Ravello environment and we have a lot data in that environment and we need to migrate all of those data to IBM Cloud Storage, think on that we created a simple solution to do everything what we need to upload all data from Ravello to IBM Cloud Storage

IBM Cloud storage is a S3 solution provider by IBM and we have a bucket to be used to upload all images from Ravello to this storage, we also have another solution to download those images from IBM Cloud Storage and import on Novello environment, but this part is out of scope of this document

Requirements

  • Ravello User

  • IBM Credentials

  • Git Hub account and commit access to novello-templates

Important
IBM Credentials
Request this information for Marcos Amorim <mamorim@redhat.com>, Patrick Rutledge <prutledge@redhat.com> or Josh Disraeli <jdisraeli@redhat.com>
Important
Git Hub Account
Send your GitHub account name to Patrick Rutledge <prutledge@redhat.com> to give your access to the novello-templates repository and configure your account to allow commit and push using SSH

Prepare your environent

Clone Ravelo2OSP Project

git clone https://github.com/redhat-gpe/ravello2osp

Install Python Modules

cd ravello2osp
pip install -r requirements.txt --user
Add your ~/.local/bin to the PATH
export PATH=$PATH:/home/USER/.local/bin
Note
 Replace USER by your username

Credentials file

cp creds.inc.example creds.inc

Edit credentials

Edit your creds.inc file and change the options below

Table 1. Credentials parameters
Variable Description

ravelloUser

Ravello username

ravelloPass

Ravello password

ravelloDomain

If your Ravello username requires the identity domain, set this value, any thing else, keep the value None

pubKeyFile

Comment this out if you want to use password authentication otherwise it will inject this password into the exporter VM in the root user

ibm_api_key

IBM API Key used to upload the images to IBM Cloud Storages. request this information!

ibm_bucket_name

IBM bucket name, set this value to gpte-novello-images

ibm_endpoint

IBM Cloud Storage endpoint, set to https://s3.us-east.cloud-object-storage.appdomain.cloud

ibm_auth_endpoint

IBM Cloud Storage auth endpoint, set to https://iam.cloud.ibm.com/identity/token

ibm_resource_id

IBM Cloud Storage resource id, request this information!

import_host

This variable set to the import server, set this value to 192.168.10.2, this value will not be used, but it will be necessary for some scripts.
Important
Request this information
Request this information for Marcos Amorim <mamorim@redhat.com>, Patrick Rutledge <prutledge@redhat.com> or Josh Disraeli <jdisraeli@redhat.com>

List of blueprints

On this spreedsheet we have a list of blueprint that we need to migrate from Ravello to IBM Cloud.

Choose a blueprint in column O, and add a note "in progress" in column V, this is necessary to avoid converting the same blueprint by more than one person, when you finish the migration, change column V, to "OK - INITIAL"

Note
INITIAL
First name and last name initials, eg: Marcos Amorim will be MA

Starting the migration

After choosing a blueprint, it’s time to start the conversion:

In the ravello2osp directory run the following command

./convert.sh BLUEPRINT_NAME
Important
BLUEPRINT_NAME
Replace BLUEPRINT_NAME by the name of blueprint that you choose on the spreadsheet
Important
You need to pay attention to some outputs on the screen
If you see an message like below: 
Wait 10 seconds before add the disks to the VM
WARNING: More than 25 disks in the blueprint, in total 30 disks.
Only 25 disks are going to be attached.

After completing the conversion, you must execute another convert-offset.sh script passing as parameter the same blueprint name

Note
 The convert.sh script adds a maximum of 25 blueprint disks, if the blueprint has more than that, we have to run the other script to continue the conversion from disk 25 to disk 49

Known Issues

Some errors are known, in case of error in the conversion, leave a comment in the spreadsheet, marking Marcos Amorim, Patrick Rutledge and Josh Disraeli with details of the error so that we can investigate.

  1. Wait 600 seconds for target connection

If you see the message bellow, go to Ravello UI and restart your application

TASK [Wait 600 seconds for target connection to become reachable/usable] **************************************************************************************************
Tuesday 25 February 2020  14:21:18 -0500 (0:00:00.099)       0:00:00.099 ******
 [WARNING]: Unhandled error in Python interpreter discovery for host exporter-exporterappmamorim-z1cavid9.srv.ravcloud.com: Failed to connect to the host via ssh: ssh:
connect to host exporter-exporterappmamorim-z1cavid9.srv.ravcloud.com port 22: Network is unreachable

The process

This how-to describes the steps required to migrate Ravello blueprints to IBM Cloud Storage, the process consists of running only one script with the parameter of the name of the blueprint you want to migrate.

The migration process consists of a few steps, as follows:

  1. Create json file from the blueprint (blueprint.json) and save it to your machine

  2. Using information from Json, creates two heat files (stack_user.yml and stack_admin.yml)

  3. Install an application on Ravello, using your Ravello username and password

  4. Create playbook to convert the disk and upload to IBM Cloud Storage

  5. Attach the blueprint disks in this application

  6. Execute the playbook to convert and upload the disks images to IBM Cloud Storage

  7. Upload the stack_user.yml, stack_admin.yaml and blueprint.json to novello-templates