elastic search snapshot management

[skm]

Hello List

I want to do the following tasks using ansible in elastic search cluster

1. create/verify s3 bucket by the name es_backup

2. create folder in s3 bucket by the name var {{ elasticsearch_cluster_id }}  - s3 module - Can we create folders using s3 module?

3. create repository using that s3 bucket and base_path (folder inside s3 bucket) option - uri module - how to implement the parameters ?

curl -XPOST localhost:9200/_snapshot/my_s3_repository -d ' 

{ 

"type": "s3", 

"settings" : { 

   "bucket": "es-backup", 

   "base_path": "logstash_cluster",

   "compress": "true",

   "access_key": "<removed>", 

   "secret_key": "<removed>" 

}

}'

Any hints/examples to help with the above steps is very welcome.

Thanks,

skm

[James Martin]

1)

You can create a bucket, take a look at the examples here: http://docs.ansible.com/s3_module.html

2)

There are no such things as "folders" in S3.  Everything is an object or bucket: http://docs.aws.amazon.com/AmazonS3/latest/UG/FolderOperations.html

3) Don't understand the last question.

- James

[skm]

Thanks!  

That cleared the confusion.

Other question I have about the snapshot management is that can the elastic search snapshots always run from only one node ? What happens when the node goes down?

What is a good practice for running snapshots as daily cron job.

Thanks,

Skm