# Prevent an ES Node from joining a cluster based on an EC2 Tag? **URL:** \ **Category:** Elasticsearch\ **Created:** [January 8, 2013, 7:45pm UTC](https://discuss.elastic.co/t/prevent-an-es-node-from-joining-a-cluster-based-on-an-ec2-tag/10268 "2013-01-08T19:45:15Z")\ **Posts on this page:** 3\ **Page:** 1
**Author:** ![tbrianjones](https://sea2.discourse-cdn.com/elastic/user_avatar/discuss.elastic.co/tbrianjones/32/1402_2.png) [@tbrianjones](https://discuss.elastic.co/u/tbrianjones)\ **Post date:** [January 8, 2013, 7:45pm UTC](https://discuss.elastic.co/t/prevent-an-es-node-from-joining-a-cluster-based-on-an-ec2-tag/10268/1 "2013-01-08T19:45:15Z")
I've got my elasticsearch.yml file set to filter ec2 nodes during discovery based on a group and a tag. I use the same security group for all my ES nodes ( jsut filtering other running servers in my account ) and thus, when I want to run a separate cluster, I assign a new value to a tag called 'index'. discovery.type: ec2 discovery.ec2.groups: elasticsearch\_node discovery.ec2.tag.index: companies My problem is that if I spin up a new Node Instance and assign a unique value to the 'index' tag during launch, this node still joins the running cluster because although the tag is different, the cluster.name in it's elasticsearch.yml file still matches the currently running cluster. It seems the tag filter only filters nodes during the discovery process ( which makes sense ), but not when a new node is looking for clusters to join. Is there a preferred way to deal with this use case? Could the AWS Gateway be configured to check the `discovery.ec2.tag.key` setting for a running cluster before a new node joins that cluster? -- ---
**Author:** ![drewr](https://sea2.discourse-cdn.com/elastic/user_avatar/discuss.elastic.co/drewr/32/7803_2.png) [@drewr](https://discuss.elastic.co/u/drewr)\ **Post date:** [January 9, 2013, 7:26pm UTC](https://discuss.elastic.co/t/prevent-an-es-node-from-joining-a-cluster-based-on-an-ec2-tag/10268/2 "2013-01-09T19:26:21Z")
Brian Jones wrote: > My problem is that if I spin up a new Node Instance and assign a > unique value to the 'index' tag during launch, this node still > joins the running cluster because although the tag is different, > the cluster.name in it's elasticsearch.yml file still matches the > currently running cluster. It seems the tag filter only filters > nodes during the discovery process ( which makes sense ), but not > when a new node is looking for clusters to join. Forgive me if I missed this from your message, but why can you not provision a different cluster name for the new nodes? Why do you want nodes that share a cluster name not to be clustered together? -Drew -- ---
**Author:** ![system](https://us1.discourse-cdn.com/elastic/original/3X/1/a/1ac57faf039f6b580b3f104ef42a2a89e41014de.png) [@system](https://discuss.elastic.co/u/system)\ **Post date:** [July 6, 2017, 2:57am UTC](https://discuss.elastic.co/t/prevent-an-es-node-from-joining-a-cluster-based-on-an-ec2-tag/10268/3 "2017-07-06T02:57:10Z")
Closed on July 6, 2017, 2:57am UTC