# Understanding ILM policy

**URL:** <https://discuss.elastic.co/t/understanding-ilm-policy/282303>\
**Category:** Logstash\
**Tags:** ilm-index-lifecycle-management\
**Created:** [August 24, 2021, 7:01am UTC](https://discuss.elastic.co/t/understanding-ilm-policy/282303 "2021-08-24T07:01:05Z")\
**Posts on this page:** 6\
**Page:** 1

<div class="post-metadata">

**Author:** ![selin](https://avatars.discourse-cdn.com/v4/letter/s/9de053/32.png) [@selin](https://discuss.elastic.co/u/selin)\
**Post date:** [August 24, 2021, 7:01am UTC](https://discuss.elastic.co/t/understanding-ilm-policy/282303/1 "2021-08-24T07:01:06Z")

</div>

Hi friends,,

It is noted i can set ILM policy in logstash attach to index

```auto
    output {
      elasticsearch {
        ilm_rollover_alias => "custom"
        ilm_pattern => "000001"
        ilm_policy => "custom_policy"
      }
    }

```

my index is currently

```auto
index => "%{[fields][myvalue]}-%{+yyyyMMdd}"

```

I want after few GB i need rollover so index wiil be

```auto
myvalue-date <- it rollover

```

what is ilm\_rollover\_alias and ilm\_pattern. How i can set it ot that behaviour?

---

<div class="post-metadata">

**Author:** ![warkolm](https://sea2.discourse-cdn.com/elastic/user_avatar/discuss.elastic.co/warkolm/32/39224_2.png) [@warkolm](https://discuss.elastic.co/u/warkolm)\
**Post date:** [August 25, 2021, 12:52am UTC](https://discuss.elastic.co/t/understanding-ilm-policy/282303/2 "2021-08-25T00:52:18Z")

</div>

It's not really efficient to roll over after a few GB. Shards should be 30-50GB in size.

> [@selin](#):
>
> ```auto
> myvalue-date <- it rollover
> 
> ```

No you cannot do that. The date in the index pattern is the date that the policy was first enacted.

---

<div class="post-metadata">

**Author:** ![selin](https://avatars.discourse-cdn.com/v4/letter/s/9de053/32.png) [@selin](https://discuss.elastic.co/u/selin)\
**Post date:** [August 25, 2021, 1:28am UTC](https://discuss.elastic.co/t/understanding-ilm-policy/282303/3 "2021-08-25T01:28:38Z")

</div>

> [@warkolm](#):
>
> It's not really efficient to roll over after a few GB. Shards should be 30-50GB in size.

wouldn't that be slow? instead of 1 gb?

---

<div class="post-metadata">

**Author:** ![warkolm](https://sea2.discourse-cdn.com/elastic/user_avatar/discuss.elastic.co/warkolm/32/39224_2.png) [@warkolm](https://discuss.elastic.co/u/warkolm)\
**Post date:** [August 25, 2021, 1:38am UTC](https://discuss.elastic.co/t/understanding-ilm-policy/282303/4 "2021-08-25T01:38:38Z")

</div>

Why would it be slow?

---

<div class="post-metadata">

**Author:** ![selin](https://avatars.discourse-cdn.com/v4/letter/s/9de053/32.png) [@selin](https://discuss.elastic.co/u/selin)\
**Post date:** [August 25, 2021, 3:24am UTC](https://discuss.elastic.co/t/understanding-ilm-policy/282303/5 "2021-08-25T03:24:40Z")

</div>

because the shared size is big, then it need each the data in a big shard rather than making it small but not too small

---

<div class="post-metadata">

**Author:** ![system](https://us1.discourse-cdn.com/elastic/original/3X/1/a/1ac57faf039f6b580b3f104ef42a2a89e41014de.png) [@system](https://discuss.elastic.co/u/system)\
**Post date:** [September 22, 2021, 4:53am UTC](https://discuss.elastic.co/t/understanding-ilm-policy/282303/7 "2021-09-22T04:53:59Z")

</div>

This topic was automatically closed 28 days after the last reply. New replies are no longer allowed.