CODEWITHSUNDEEP
ansible
Tarasovych
Tarasovych

Reputation: 2398

Can not sync two S3 buckets using ansible s3_sync module

What do I have: two S3 buckets named "s3_sync-ansible-test-1" and "s3_sync-ansible-test-2".

Buckets structure:

s3_sync-ansible-test-1
|-folder_1
  |-file_1.txt
  |-file_2.txt
|-folder_2
  |-file_2.txt

s3_sync-ansible-test-2 bucket is empty.

What do I want to do: sync everything between s3_sync-ansible-test-1 and s3_sync-ansible-test-2.

What did I tried:

    - name: "Sync S3 buckets"
      register: sync_s3
      s3_sync:
        bucket: s3_sync-ansible-test-2
        file_root: /s3_sync-ansible-test-1

    - debug:
        msg: "{{ sync_s3 }}"

Debug output is

    "msg": {
        "changed": false, 
        "failed": false, 
        "filelist_actionable": [], 
        "filelist_initial": [], 
        "filelist_local_etag": [], 
        "filelist_s3": [], 
        "filelist_typed": [], 
        "uploads": []
    }

At the same time, manual command

aws s3 sync s3://s3_sync-ansible-test-1 s3://s3_sync-ansible-test-2

works as expected.

I've also tried to change file_root to s3://s3_sync-ansible-test-2 without success.

What could be wrong?

Upvotes: 0

Views: 378

Answers (1)

Nagle Zhang
Nagle Zhang

Reputation: 74

As the document of s3_sync described:

file_root: / required File/directory path for synchronization.This is a local path. This root path is scrubbed from the key name, so subdirectories will remain as keys

the file_root is a local path. which means it can't be a s3 bucket.

if you really want to using this module , you can use s3fs mount s3 bucket to your instance, then using ansible s3_sync to sync it.

Upvotes: 1

Related Questions