0969d53c6c
This patch adds support for upgrading Cassandra datastore - both single instance cluster and multiple-node cluster. It is achievied in a similar way to MySQL/Mariadb. The key difference is draining nodes before upgrade and upgrading sstables after succesful volume mounting operation. Cassandra upgrade steps have been taken from https://docs.datastax.com/en/upgrade/doc/upgrade/ddac/upgdDDACMinor51x.html In case of trove it looks as follows: For each node do: + enter restart mode + drain and stop node + preserve configuration files and home directory + unmount volume + create instance with new datastore version + mount volume + restore configuration and home directory + start Cassandra + upgrade sstables Because we need to upgrade seeds first there is an additional parameter in rolling_upgrade_cluster function - a function which is used to so sort nodes in proper order. Story: #2005410 Task: #30426 Co-Authored-By: Przemysław Godek <p.godek@partner.samsung.com> Change-Id: I9f25606d51b35e264a8c36f9524ae57b0a5780d0 Signed-off-by: Kasper Hasior <k.hasior@samsung.com>
Trove
Trove is Database as a Service for OpenStack.
Getting Started
If you'd like to run from the master branch, you can clone the git repo:
For information on how to contribute to trove, please see CONTRIBUTING.rst and HACKING.rst
You can raise bugs here: Bug Tracker
The plan for trove can be found at Trove Specs
- Release notes for the project can be found at:
Python client
Python-troveclient is a client for Trove.
Dashboard plugin
Trove-dashboard is OpenStack dashbaord plugin for Trove.