Cluster Admin

Question 1

How do you check the status of the nodes?

Answer 1

GET _cat/nodes?v

Question 2

How do you check the status of indices?

Answer 2

GET _cat/indices/v

Question 3

How do you see on which nodes the shards of an index is allocated?

Answer 3

GET _cat/shards/< index >?v

Question 4

How does elastic determine in which nodes to put shards?

Answer 4

Elastic will make its best judment based on the settings of the index, unless you do allocation filtering.

Question 5

How do you specify the number of replicas and shards of an index?

Answer 5

In the index settings:

PUTS < index >
{
   settings: {
      number_of_replicas: X,
      number_of_shards: Y.
   }
}

Question 6

What’s a shortcut to modify only the index settings?

Answer 6

PUT < index >/_settings

Question 7

How to override index routing allocation (filtering)?

Answer 7

IN the index settings:

# To exclude a node
{
  "index.routing.allocation.exclude._name": "< node name >, ...."
}

# To revert the exclusion:
{
  "index.routing.allocation.exclude._name":  null
}

TODO: Check the api doc to other variables and use cases.

Question 8

What are some fail safe measurement elastic has in place for shard allocation?

Answer 8

Elastic will only reallocate the shards if the index will remain in a green state.

Question 9

How to do cluster routing allocation?

Answer 9

PUT _cluster/settings
{
“transient”: {
“cluster.routing.allocation.exclude._name”: “< node name >, …”

Question 10

What is transient and persistent cluster settings?

Answer 10

Transient won’t survive a cluster restart.

Question 11

What is routing allocation in elastic and what are some of the use cases?

Answer 11

Allows you to manually create rules that dictate where shards are allocated.

A common use case is taking a node down from the cluster, so you can filter that single node, wait for the cluster to rebalance and then shut it down.

Question 12

How do you see the node attributes?

Answer 12

GET _cat/nodeattrs?v

Question 13

What is cluster routing allocation awareness and what are some of the use cases?

Answer 13

Allows to create some allocation rules for where to put primaries/replicas.

A tipical use case is avoiding putting all data into a single zone. So you can create “zone” custom attribute and use that in the allocation.

Question 14

What are the 2 ways to setup rounting allocation affinity?

Answer 14

Either put in the config.yml or use the API.

Question 15

How do you setup cluster routing awareness?

Answer 15

PUT _cluster/settings
{
   "persistent": {
       "cluster.routing.allocation.awareness.attributes": "< attr name >"
}

Question 16

What is force awareness and what is it used for and how do you set it up?

Answer 16

In a situation where you might not have enough space to re-allocate of the shards when a whole zone goes down, you might want to prevent those shards being re-allocated in case a whole zone goes down.

PUT _cluster/settings
{
   "persistent": {
       "cluster.routing.allocation.awareness.force.< attr name >.values: " < list of all possible values for the attribute > "
}

Question 17

What is the cluster allocation API?

Answer 17

Explains why your cluster is allocated a certain way.

GET _cluster/allocation/explain

Question 18

How do you setup filesystem based backup on elastic?

Answer 18

• Whitelist the repository where the backup is gonna be saved

In config/elasticsearch.yml:

path. repo: /home/elastic/snaphosts
- Make sure the directory exists
- Restart elastic

Question 19

How do you create a Filesystem snapshot repository with the API?

Answer 19

PUT _snapshot/< repo name >
{
   type: "fs",
   settings: {
       location: "/home/elastic/snapshots",
}

PS: This doesn’t save the data, just creates a “repository”.

Question 20

How do you create a backup of an index?

Answer 20

• With the repository previously created:

PUT _snapshot/< repo name>/< snapshot name >
{
indices: “< index name >”
}

Question 21

How do I wait for an API call to finish for async calls?

Answer 21

Add “?wait_for_completion=true” to the call.

Question 22

How do you restore a backup snapshot?

Answer 22

POST _snapshot/< repo name >/< snapshot name >/_restore
{
indices: “< the index to restore >”
}

Question 23

How do you backup all indices that start with a certain pattern?

Answer 23

You can use wildcards in the index name for the snapshot API.

Question 24

How do rename indices when restoring a backup?

Answer 24

Use the “rename_pattern” option with a regular expression.

{
rename_pattern: “(.+)”,
rename_replacement: “$1_restored”
}

Question 25

How do you create a tiered architecture (hot/warm/cold) in elastic?

Answer 25

• Create a custom attribute named, for example “temp” in the config files.
• In the index settings, configure the routing allocation (index.routing.allocation.require.temp: “hot/warm”).

Question 26

How do you configure a remote cluster to your cluster?

Answer 26

In the cluster settings:

PUT _cluster/settings
{
   "persistent": {
       "cluster": {
           "remote": {
                 "< cluster name >": {
                    seeds: ["< ip >:9300"]
}

Make sure to use the transport port (9300).

Question 27

What protocol/port is used for search in between clusters?

Answer 27

The elastic “transport” protocol (port 9300).

Question 28

How do you search on a remote cluster?

Answer 28

GET < remote name >: < index name >/_search