Jonathan Hurley created AMBARI-12228:
----------------------------------------
Summary: Oozie Rolling Upgrade Orchestration Is Incorrect
Key: AMBARI-12228
URL: https://issues.apache.org/jira/browse/AMBARI-12228
Project: Ambari
Issue Type: Bug
Components: ambari-server
Affects Versions: 2.0.0
Reporter: Jonathan Hurley
Assignee: Jonathan Hurley
Priority: Critical
Fix For: 2.1.0
Here is an example of the new orchestration for an upgrade from 2.2 to 2.3
- Before Upgrade Begins
-# Execute a custom script to stop all oozie servers, one at a time
-# Execute the database upgrade using the 2.3 bits
-# Execute the sharelib creation and upload using the 2.3 bits
- For each Oozie server, which is currently down
-# backup the oozie configurations
-# conf-select and hdp-select
-# restore configurations
-# Prepare /usr/hdp/current/oozie/libext-customer and gets ready to upgrade the
WAR
-# \{oozie_setup_sh\} prepare-war \{oozie_secure\} -d
\{oozie_libext_customer_dir\}
-# Start the Oozie server
On a downgrade, we will perform a similar series of steps, except a few
differences:
- No database upgrade will be performed
- The bits used to generate the sharelib are going to be the downgrade stacks,
ie 2.2
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
