** version 2.2.0**
Describe the issue:
I added 2 new data nodes on in my opensearch Cluster to store the logs of our API-Gateway
I dont want .opensearch* and .opendistro* indexes to be created on these 2 new nodes
So I set a rack id for my existing nodes called applicative_rack
And i created ( successfully) 2 index templates for these indexes as follow : ( to force allocating these system indexes on the applicative_rack and not on the new nodes for which rack id is different ) :
POST /_index_template/opendistro { “index_patterns”: [“.opendistro*”], “template”: { “settings”: { “index.routing.allocation.require.rack”: “applicative_rack” } }}
POST /_index_template/opensearch { “index_patterns”: [“.opensearch*”], “template”: { “settings”: { “index.routing.allocation.require.rack”: “applicative_rack” } }}
But when I started my new Opensearch data nodes :
it seems the .opendistro* and .opensearch* indexes are created on them ( the new nodes ) and the index templates are absolutely NOT applied ( issue occurs only for these system indexes, no issue with any others indexes… )
Why can t I force index allocation for : .opendistro* and .opensearch* indexes ?
( mainly:
.opendistro-job-scheduler-lock
.opendistro-ism-managed-index-history
.opendistro-alerting-config
.opensearch-notifications-config
)
Is there any other way to proceed and force allocation for these system indexes ?
Regards