---
id: c98fd83e-6481-11ee-8c99-0242ac120002
---

# Unassigned Shards Troubleshooting in Elasticsearch
---

This incident type refers to situations where some shards in an Elasticsearch cluster have not been assigned to any node. This can happen due to various reasons such as node failure, network issues, and configuration problems. Unassigned shards can lead to degraded performance and data loss. Troubleshooting unassigned shards involves identifying the root cause of the problem and taking appropriate actions to resolve it.

### Parameters
```shell
export ELASTICSEARCH_HOST="PLACEHOLDER"

export ELASTICSEARCH_PORT="PLACEHOLDER"

export NODE_ID="PLACEHOLDER"

export CLUSTER_NAME="PLACEHOLDER"

export NODE_NAME="PLACEHOLDER"

export NUMBER_OF_SECONDS="PLACEHOLDER"
```

## Debug

### Check Elasticsearch cluster health status
```shell
curl -XGET ${ELASTICSEARCH_HOST}:${ELASTICSEARCH_PORT}/_cluster/health
```

### Get a list of all unassigned shards
```shell
curl -XGET ${ELASTICSEARCH_HOST}:${ELASTICSEARCH_PORT}/_cat/shards | grep UNASSIGNED
```

### Get the reason for unassigned shards
```shell
curl -XGET ${ELASTICSEARCH_HOST}:${ELASTICSEARCH_PORT}/_cluster/allocation/explain?pretty
```

### Check Elasticsearch cluster and node settings
```shell
curl -XGET ${ELASTICSEARCH_HOST}:${ELASTICSEARCH_PORT}/_cluster/settings

curl -XGET ${ELASTICSEARCH_HOST}:${ELASTICSEARCH_PORT}/_nodes/${NODE_ID}/settings
```

### Check Elasticsearch cluster and node logs for errors
```shell
tail -f /var/log/elasticsearch/${CLUSTER_NAME}.log

tail -f /var/log/elasticsearch/${CLUSTER_NAME}/${NODE_NAME}.log
```

### Get Elasticsearch configuration settings
```shell
curl -XGET ${ELASTICSEARCH_HOST}:${ELASTICSEARCH_PORT}/_cluster/settings?include_defaults
```

## Repair

### Restarting Elasticsearch nodes: Restarting Elasticsearch nodes can sometimes resolve issues related to unassigned shards. This can be done one node at a time to minimize the impact on the cluster.
```shell


#!/bin/bash



# Stop Elasticsearch service

sudo systemctl stop elasticsearch.service



# Wait for Elasticsearch to stop

sleep ${NUMBER_OF_SECONDS}



# Start Elasticsearch service

sudo systemctl start elasticsearch.service



# Wait for Elasticsearch to start

sleep ${NUMBER_OF_SECONDS}



# Check Elasticsearch status

sudo systemctl status elasticsearch.service


```


This incident type refers to situations where some shards in an Elasticsearch cluster have not been assigned to any node. This can happen due to various reasons such as node failure, network issues, and configuration problems. Unassigned shards can lead to degraded performance and data loss. Troubleshooting unassigned shards involves identifying the root cause of the problem and taking appropriate actions to resolve it.


```shell
export ELASTICSEARCH_HOST="PLACEHOLDER"

export ELASTICSEARCH_PORT="PLACEHOLDER"

export NODE_ID="PLACEHOLDER"

export CLUSTER_NAME="PLACEHOLDER"

export NODE_NAME="PLACEHOLDER"

export NUMBER_OF_SECONDS="PLACEHOLDER"
```


### Check Elasticsearch cluster health status

```shell
curl -XGET ${ELASTICSEARCH_HOST}:${ELASTICSEARCH_PORT}/_cluster/health
```

### Get a list of all unassigned shards

```shell
curl -XGET ${ELASTICSEARCH_HOST}:${ELASTICSEARCH_PORT}/_cat/shards | grep UNASSIGNED
```

### Get the reason for unassigned shards

```shell
curl -XGET ${ELASTICSEARCH_HOST}:${ELASTICSEARCH_PORT}/_cluster/allocation/explain?pretty
```

### Check Elasticsearch cluster and node settings

```shell
curl -XGET ${ELASTICSEARCH_HOST}:${ELASTICSEARCH_PORT}/_cluster/settings

curl -XGET ${ELASTICSEARCH_HOST}:${ELASTICSEARCH_PORT}/_nodes/${NODE_ID}/settings
```

### Check Elasticsearch cluster and node logs for errors

```shell
tail -f /var/log/elasticsearch/${CLUSTER_NAME}.log

tail -f /var/log/elasticsearch/${CLUSTER_NAME}/${NODE_NAME}.log
```

### Get Elasticsearch configuration settings

```shell
curl -XGET ${ELASTICSEARCH_HOST}:${ELASTICSEARCH_PORT}/_cluster/settings?include_defaults
```


### Restarting Elasticsearch nodes: Restarting Elasticsearch nodes can sometimes resolve issues related to unassigned shards. This can be done one node at a time to minimize the impact on the cluster.

```shell


#!/bin/bash



# Stop Elasticsearch service

sudo systemctl stop elasticsearch.service



# Wait for Elasticsearch to stop

sleep ${NUMBER_OF_SECONDS}



# Start Elasticsearch service

sudo systemctl start elasticsearch.service



# Wait for Elasticsearch to start

sleep ${NUMBER_OF_SECONDS}



# Check Elasticsearch status

sudo systemctl status elasticsearch.service


```


Unassigned Shards Troubleshooting in Elasticsearch

Overview

Parameters

Debug

Check Elasticsearch cluster health status

Get a list of all unassigned shards

Get the reason for unassigned shards

Check Elasticsearch cluster and node settings

Check Elasticsearch cluster and node logs for errors

Get Elasticsearch configuration settings

Repair

Learn more

Related Runbooks

Support