# ILM policy not triggering a new index

**URL:** <https://discuss.elastic.co/t/ilm-policy-not-triggering-a-new-index/376066>\
**Category:** Elasticsearch\
**Tags:** ilm-index-lifecycle-management\
**Created:** [March 18, 2025, 11:34am UTC](https://discuss.elastic.co/t/ilm-policy-not-triggering-a-new-index/376066 "2025-03-18T11:34:44Z")\
**Posts on this page:** 2\
**Page:** 1

<div class="post-metadata">

**Author:** ![aravinds257](https://avatars.discourse-cdn.com/v4/letter/a/7993a0/32.png) [@aravinds257](https://discuss.elastic.co/u/aravinds257)\
**Post date:** [March 18, 2025, 11:34am UTC](https://discuss.elastic.co/t/ilm-policy-not-triggering-a-new-index/376066/1 "2025-03-18T11:34:44Z")

</div>

I have index created on a weekly basis via a pipeline which creates a index ending with \_yyyymmddw, but was relying on ILM policy sharding based on if it reaches the 50gb limit. Even though the index action shows as rollover nothing is happening and it has reached almost 600gb.

I see the error of index rollover alias. is there anyway it can pick up the first index name it creates for the rollover?

ILM policy on the index

```auto
{
  "policy": "1-week-retention",
  "phase_definition": {
    "min_age": "0ms",
    "actions": {
      "rollover": {
        "max_age": "30d",
        "max_primary_shard_size": "50gb"
      },
      "set_priority": {
        "priority": 100
      }
    }
  },
  "version": 3,
  "modified_date_in_millis": 1742225274621
}

```

ILM explain

```auto
{
  "indices": {
    "nifiaudit_60187_review_20250312": {
      "index": "nifiaudit_60187_review_20250312",
      "managed": true,
      "policy": "1-week-retention",
      "index_creation_date_millis": 1742225294227,
      "time_since_index_creation": "20.12h",
      "lifecycle_date_millis": 1742225294227,
      "age": "20.12h",
      "phase": "hot",
      "phase_time_millis": 1742296780127,
      "action": "rollover",
      "action_time_millis": 1742288451847,
      "step": "ERROR",
      "step_time_millis": 1742297380160,
      "failed_step": "check-rollover-ready",
      "is_auto_retryable_error": true,
      "failed_step_retry_count": 7,
      "step_info": {
        "type": "illegal_argument_exception",
        "reason": "setting [index.lifecycle.rollover_alias] for index [nifiaudit_60187_review_20250312] is empty or not defined"
      },
      "phase_execution": {
        "policy": "1-week-retention",
        "phase_definition": {
          "min_age": "0ms",
          "actions": {
            "rollover": {
              "max_primary_shard_size": "50gb"
            },
            "set_priority": {
              "priority": 100
            }
          }
        },
        "version": 4,
        "modified_date_in_millis": 1742288359047
      }
    }
  }
}

```

---

<div class="post-metadata">

**Author:** ![GinkoLucas](https://sea2.discourse-cdn.com/elastic/user_avatar/discuss.elastic.co/ginkolucas/32/107055_2.png) [@GinkoLucas](https://discuss.elastic.co/u/GinkoLucas)\
**Post date:** [March 25, 2025, 3:25pm UTC](https://discuss.elastic.co/t/ilm-policy-not-triggering-a-new-index/376066/2 "2025-03-25T15:25:24Z")

</div>

Hi,  
ILM (Index Lifecycle Management) works with aliases, and according to your error, you don't have one.  
The documentation for Rollover and how it works with Aliases : [Rollover | Elasticsearch Guide [8.17] | Elastic](https://www.elastic.co/guide/en/elasticsearch/reference/current/ilm-rollover.html)

Do you know about data streams? In your case, it might be interesting, it simplifies the system I find. Are your data timestamped? Do you need to modify them, or just index them?