# Elastic shards and replica
**URL:** https://discuss.elastic.co/t/elastic-shards-and-replica/272552
**Category:** Elasticsearch
**Tags:** ilm-index-lifecycle-management
**Created:** [May 10, 2021, 8:25am UTC](https://discuss.elastic.co/t/elastic-shards-and-replica/272552 "2021-05-10T08:25:38Z")
**Posts on this page:** 7
**Page:** 1
### Author:  [@TheHunter1](https://discuss.elastic.co/u/TheHunter1)
#### Post date: [May 10, 2021, 8:25am UTC](https://discuss.elastic.co/t/elastic-shards-and-replica/272552/1 "2021-05-10T08:25:38Z")
Hello everyone,
I would like to understand something about shards and replica,
I have a cluster of 3 DATA nodes (2 Hot and 1 Warm), I have configured ILM to have one shard and one replica in the HOT nodes, and then once it passed to the WARM node, I delete the replica and just save the shards in that node.
Today when I restarted one of my HOT nodes, the stat of some of shards in the `WARM` nodes became `YELLOW`
I would like to understand why these shards have been affected as they are in the WARM node and not in the Hot nodes ?
wehn I run `GET _cluster/allocation/explain?pretty` I get some output like these :
```auto
"allocate_explanation" : "cannot allocate because allocation is not permitted to any of the nodes",
"explanation" : "index has a preference for tiers [data_warm,data_hot] and node does not meet the required [data_warm] tier"
"explanation" : "index has a preference for tiers [data_warm,data_hot] and node does not meet the required [data_warm] tier"
"explanation" : "a copy of this shard is already allocated to this node [[metricbeat-7.10.1-2021.03.19-000003][0], node[nN9JxEJrTQ6wIBoQLHC-WA], [P], s[STARTED], a[id=MEt6XzEhS1G2WH6n-2-1jg]]"
```
Thanks for your help
---
### Author:  [@TheHunter1](https://discuss.elastic.co/u/TheHunter1)
#### Post date: [May 10, 2021, 9:36am UTC](https://discuss.elastic.co/t/elastic-shards-and-replica/272552/2 "2021-05-10T09:36:28Z")
UPDATE: I restarted all the nodes again, and now the cluster health is `GREEN`
But It would be cool if someone can explain to me why the shards in the `WARM` node had been affected when I restarted the `HOT` node !
Thanks 🙂
---
### Author:  [@DavidTurner](https://discuss.elastic.co/u/DavidTurner)
#### Post date: [May 10, 2021, 10:03am UTC](https://discuss.elastic.co/t/elastic-shards-and-replica/272552/3 "2021-05-10T10:03:02Z")
> [@TheHunter1](#):
>
> once it passed to the WARM node, I delete the replica
The allocation explain output indicates that you did not delete the replica on this index even though it was supposed to be in the warm phase.
---
### Author:  [@TheHunter1](https://discuss.elastic.co/u/TheHunter1)
#### Post date: [May 10, 2021, 11:16am UTC](https://discuss.elastic.co/t/elastic-shards-and-replica/272552/4 "2021-05-10T11:16:43Z")
Here is how I Configured ILM:
This should delete all the replicas in the Warm phase, isn't it ?
---
### Author:  [@DavidTurner](https://discuss.elastic.co/u/DavidTurner)
#### Post date: [May 10, 2021, 1:08pm UTC](https://discuss.elastic.co/t/elastic-shards-and-replica/272552/5 "2021-05-10T13:08:36Z")
> [@TheHunter1](#):
>
> This should delete all the replicas in the Warm phase, isn't it ?
Looks like it should, but apparently it didn't have that effect. Are you sure this policy applies to the index that caused you trouble?
---
### Author:  [@TheHunter1](https://discuss.elastic.co/u/TheHunter1)
#### Post date: [May 10, 2021, 1:21pm UTC](https://discuss.elastic.co/t/elastic-shards-and-replica/272552/6 "2021-05-10T13:21:43Z")
I have configured all my policies the same way.
I will investigate more to find out if there is something wrong in the cluster, and then I will restartd one hot node, to see if I will get the same result as the first time
---
### Author:  [@system](https://discuss.elastic.co/u/system)
#### Post date: [June 7, 2021, 1:21pm UTC](https://discuss.elastic.co/t/elastic-shards-and-replica/272552/7 "2021-06-07T13:21:58Z")
This topic was automatically closed 28 days after the last reply. New replies are no longer allowed.