We have developed an application based on Elasticsearch. The data held is currently relatively small and the user interaction is very low. The application was first developed a few years ago based on Elasticsearch 6.2.4. There are currently only 3 distinct deployments (totally independent - no data is shared and no connection between the deployments).
We are developing release 2 and obviously want to use the latest Elasticsearch version (8.x).
As the data involved is small and the usage means that an upgrade slot can easily be arranged with no users on-line.
The documentation is quite confusing, some articles implying that that some intermediary release must first be taken and other articles saying that it can be done in one step with re-indexing.
To upgrade from 6.2.4 to 8.1 you will need to perform at least 3 upgrades.
First you will need to go to the last 6.X version, which is 6.8, this version have the upgrade assistant that will tell you what you need to change to be able to upgrade to 7.X.
You need this step because there are a couple of breaking changes from 6.X to 7.X.
Once on 6.8, you can upgrade to 7.17 which is the last release of the 7.x branch and also has an upgrade assistant to help you upgrade to 8.1, again, thera are breaking changes from 7.X to 8.X.
Since you said that the data is currently small and barely used, maybe a reindex will help.
In this case a reindex would not be an upgrade, you would need to create a completely new cluster running on version 8.1 and then reindex your data from the 6.2.4 cluster usgin the Reindex API or Logstash.
Thank you very much, excellent information.
Please bear with me I am a novice with Elasticsearch.
I appreciate that on our development system we will have to go through three updates and ensure that the code we have written to exploit Elasticsearch still functions correctly.
Once we have a releasable product, do we have to go through the same 3 steps with each customer system or can we use the new cluster/reindex method?
Sorry I am still getting to grips with the concepts involved
Apache, Apache Lucene, Apache Hadoop, Hadoop, HDFS and the yellow elephant
logo are trademarks of the
Apache Software Foundation
in the United States and/or other countries.