Hi, I'm facing some problem in Index life cycle policy in which indexes are not roll overing and delete phase is not working. Is there is a best practice to do so?
It would help if you showed your policy as well as indices stats.
How are you sending data to the index? Are you writing to the write alias?
What does your Metricbeat config look like?
Here the metricbeat-7.6.2-000001 is the index pattern name.
That is not the write alias so is the reason it is not working.
Can you please help me out with the right one
Have a look at this guide to setting up an ILM policy. When you create the first index you specify a write index alias (test-alias in the example in the docs I linked to). This is what you configure Logstash and/or Metricbeat to write to, not the actual index name. This will allow the alias to move to a new index once rollover happens.
Thankyou

Here the data is pushing through the logstash by using index => "%{[@metadata][beat]}-%{[@metadata][version]}-000001". As it is passing through the logstash, rollovering is happening but it is not deleting the old index and it is keep on adding the size of the old index. Whether I need to change anything in this config. file.


