# Upgrading Elasticsearch 8.19.2 to 9.1.2 **URL:** \ **Category:** Elastic Search\ **Created:** [August 27, 2025, 6:56pm UTC](https://discuss.elastic.co/t/upgrading-elasticsearch-8-19-2-to-9-1-2/381391 "2025-08-27T18:56:55Z")\ **Posts on this page:** 2\ **Page:** 1
**Author:** ![mst3r25](https://sea2.discourse-cdn.com/elastic/user_avatar/discuss.elastic.co/mst3r25/32/97109_2.png) [@mst3r25](https://discuss.elastic.co/u/mst3r25)\ **Post date:** [August 27, 2025, 6:56pm UTC](https://discuss.elastic.co/t/upgrading-elasticsearch-8-19-2-to-9-1-2/381391/1 "2025-08-27T18:56:55Z")
I am having a nightmare upgrading from 8.19.2 to 9.1.2. Self-host Air-gap netowrk 7 node cluster RHEL 8.10 Java 17 1. Upgrade advisory says I am good. 1. I have two warnings one for Elasticsearch and one for Kibana. 1. ES: xpack.monitoring.collection.enabled deprecated 2. Kibana: ILM policy deprecated. 2. I remove on node from the cluster. rpm -Uvh and it upgrades fine. 3. When I go to going the upgrade node back in the cluster I get an error saying various indices from es-7.17 is present[examples: .kabana-7.17.1-001, .security-tokens-7….]. I say various because I get a different message per restart. 4. Is there a way to flush or delete these old indices or migrate them so they are compadable with es 9.X? 5. I’ve upgrade and downgrade four times now and I don’t want to keep goign through this cycle. ---
**Author:** ![Keith\_Massey](https://sea2.discourse-cdn.com/elastic/user_avatar/discuss.elastic.co/keith_massey/32/83666_2.png) [@Keith\_Massey](https://discuss.elastic.co/u/Keith_Massey)\ **Post date:** [August 28, 2025, 2:16pm UTC](https://discuss.elastic.co/t/upgrading-elasticsearch-8-19-2-to-9-1-2/381391/2 "2025-08-28T14:16:17Z")
Did you click on the “Migrate System Indices” button in the upgrade assistant? It can be easy to miss, and I think it would cause problems like this.