I am running elasticsearch1.1.1. I am planning to upgrade our test and production cluster to elasticsearch 2.0.
I would like to know the steps if I want to revert back to 1.1.1 if the upgrade does not go well.
Thanks
Malini
You cannot due to underlying Lucene changes.
Make sure you take a snapshot before an upgrade so you can restore this.
I have data and logs folders outside the elasticsearch installation directory (/clusterdata/data and /clusterdata/logs).
Does the following steps sound ok?
Take the snapshot of the data.
Unzip elasticsearch2.0 and modify the elasticsearch.yml so that the data and logs paths are pointing to /clusterdata/data and /clusterdata/logs.
Shut down all the elasticssearch 1.1.1 servers
Start elasticsearch 2.0 one node at a time.
If the upgrade does not go well, shut down 2.0 and restart 1.1.1 and restore data from the snapshot?
I don't think that will work, you will need to move the data aside completely before moving back to 1.1.1.
What do you mean by moving data aside?
If upgrade does not work, should we delete the data folder and restart 1.1.1 and restore data from snapshot?
Yes, that will work as well.