Replica management in Elasticsearch - Cheat Sheet & Quick Revision

Recall & Review

beginner

What is a replica in Elasticsearch?

A replica is a copy of a primary shard in Elasticsearch. It provides fault tolerance and improves search performance by distributing the load.

beginner

How do replicas improve Elasticsearch availability?

Replicas allow Elasticsearch to continue serving data even if a primary shard fails, because the replica shard can take over without data loss.

intermediate

How can you change the number of replicas for an existing Elasticsearch index?

You can update the number of replicas using the Update Index Settings API with a command like:
PUT /my_index/_settings { "index": { "number_of_replicas": 2 } }

intermediate

What happens if you set the number of replicas to 0 in Elasticsearch?

Setting replicas to 0 means there are no copies of primary shards. This reduces fault tolerance and search performance but saves disk space.

intermediate

Can replicas be allocated on the same node as their primary shards?

No. Elasticsearch ensures replicas are allocated on different nodes than their primary shards to avoid data loss if a node fails.

What is the main purpose of replicas in Elasticsearch?

A To provide fault tolerance and improve search speed

B To store backup data on the same node

C To reduce the number of shards

D To increase indexing speed only

Replicas provide copies of primary shards on different nodes to protect against failure and improve search performance.

How do you change the number of replicas for an index?

A Using the Update Index Settings API

B By deleting and recreating the index

C By changing the cluster settings only

D By restarting Elasticsearch

You update the number of replicas with the Update Index Settings API without needing to recreate the index.

If a primary shard fails, what happens to its replica shard?

A It is deleted automatically

B It remains inactive

C It becomes the new primary shard

D It merges with other shards

The replica shard is promoted to primary to keep the data available.

Can replicas be stored on the same node as their primary shards?

A Yes, always on the same node

B No, replicas are stored on different nodes

C Only if manually configured

D Only for small clusters

Elasticsearch automatically places replicas on different nodes to avoid data loss.

What is a consequence of setting number_of_replicas to 0?

A Indexing speed decreases

B More disk space is used

C Search performance improves

D No replicas exist, reducing fault tolerance

With zero replicas, there are no copies of data, so if a primary shard fails, data is unavailable.

Explain what replicas are in Elasticsearch and why they are important.

Think about how Elasticsearch keeps data safe and fast to search.

Describe how to change the number of replicas for an existing index and what effect it has.

Consider the command to update settings and what happens after.

Practice

1.

What is the main purpose of setting replicas in an Elasticsearch index?

easy

A. To encrypt data for security

B. To delete old data automatically

C. To compress data for storage savings

D. To create copies of data for faster search and fault tolerance

Solution

1. Step 1: Understand replica role

   Replicas are copies of the original data that help improve search speed and provide backup in case of failure.

2. Step 2: Compare options

   Only To create copies of data for faster search and fault tolerance correctly describes replicas as copies for speed and safety; others describe unrelated features.

3. Final Answer:

   To create copies of data for faster search and fault tolerance -> Option D

4. Quick Check:

   Replicas = copies for speed and safety [OK]

Hint: Replicas are copies that speed up search and protect data [OK]

Common Mistakes:

• Confusing replicas with data deletion
• Thinking replicas compress data
• Assuming replicas encrypt data

2.

Which of the following is the correct syntax to update the number of replicas to 2 for an existing index named my_index using Elasticsearch REST API?

easy

A. POST /my_index/_settings { "number_of_replicas": 2 }

B. PUT /my_index/_settings { "number_of_replicas": 2 }

C. GET /my_index/_settings { "number_of_replicas": 2 }

D. DELETE /my_index/_settings { "number_of_replicas": 2 }

Solution

1. Step 1: Identify correct HTTP method for updating settings

   Elasticsearch uses PUT to update index settings like replicas.

2. Step 2: Match syntax with method

   PUT with the path /my_index/_settings and JSON body setting number_of_replicas to 2 is correct.

3. Final Answer:

   PUT /my_index/_settings { "number_of_replicas": 2 } -> Option B

4. Quick Check:

   Update settings uses PUT method [OK]

Hint: Use PUT to update index settings like replicas [OK]

Common Mistakes:

• Using POST instead of PUT for settings update
• Using GET which only retrieves settings
• Using DELETE which removes resources

3.

Given an index products with number_of_replicas set to 1, what will be the total number of shards (primary + replicas) if the index has 3 primary shards?

medium

A. 6

B. 4

C. 9

D. 3

Solution

1. Step 1: Understand shards and replicas

   Each primary shard has replicas equal to number_of_replicas. Total shards = primary shards + replicas.

2. Step 2: Calculate total shards

   3 primary shards + 1 replica each = 3 + 3 = 6 total shards.

3. Final Answer:

   6 -> Option A

4. Quick Check:

   Total shards = primary + replicas = 3 + 3 = 6 [OK]

Hint: Total shards = primary shards x (1 + replicas) [OK]

Common Mistakes:

• Counting only primary shards
• Adding replicas as 1 total instead of per shard
• Multiplying incorrectly

4.

What is wrong with this Elasticsearch index settings update request to set replicas to 3?

PUT /store/_settings
{
  "number_of_replicas": "3"
}

medium

A. The index name should be in quotes

B. PUT method cannot be used to update settings

C. The number_of_replicas value should be an integer, not a string

D. The JSON body is missing the settings wrapper

Solution

1. Step 1: Check data type of number_of_replicas

   number_of_replicas must be an integer, but here it is given as a string "3".

2. Step 2: Validate other parts

   PUT is correct method, index name does not require quotes in URL, and settings wrapper is optional in this context.

3. Final Answer:

   The number_of_replicas value should be an integer, not a string -> Option C

4. Quick Check:

   number_of_replicas must be integer [OK]

Hint: Use integer values for number_of_replicas, not strings [OK]

Common Mistakes:

• Putting number_of_replicas as a string
• Thinking PUT is wrong method
• Adding unnecessary JSON wrappers

5.

You want to increase the number of replicas for an index logs from 1 to 2 without downtime. Which approach is correct?

1. Update number_of_replicas setting to 2 using the REST API.
2. Reindex all data into a new index with 2 replicas.
3. Delete the index and recreate it with 2 replicas.
4. Change the number of primary shards to 2.

hard

A. Only step 1 is correct

B. Only step 2 is correct

C. Steps 1 and 2 are correct

D. Steps 3 and 4 are correct

Solution

1. Step 1: Understand replica update without downtime

   Elasticsearch allows changing number_of_replicas dynamically without downtime by updating settings.

2. Step 2: Evaluate other steps

   Reindexing or deleting index causes downtime; changing primary shards is unrelated to replicas.

3. Final Answer:

   Only step 1 is correct -> Option A

4. Quick Check:

   Update replicas via settings without downtime [OK]

Hint: Change replicas via settings update to avoid downtime [OK]

Common Mistakes:

• Thinking reindexing is needed to change replicas
• Deleting index causes data loss and downtime
• Confusing primary shards with replicas