Hello,
This is kind of a general question that I didnt manage to find it reply reading some of the PR and comments in github.
I am running:
]# rclone version
rclone v1.55.0-DEV
- os/type: linux
- os/arch: amd64
- go/version: go1.14.12
- go/linking: dynamic
- go/tags: none
]# lsb_release -a
LSB Version: :core-4.1-amd64:core-4.1-noarch
Distributor ID: CentOS
Description: CentOS Linux release 8.3.2011
Release: 8.3.2011
Codename: n/a
To copy content in a file system to Ceph S3 backend. My rclone.conf looks like:
[s3prevessin]
type = s3
provider = Ceph
env_auth = false
access_key_id = XXXXX
secret_access_key = XXXX
endpoint = host
acl = public-read
And the command is simply:
rclone --checksum -P --s3-upload-cutoff=1G --s3-chunk-size=256M --transfers=2 --checkers=4 --log-level DEBUG copy /mnt/media_share/media_data s3prevessin:publicqa/mnt/media_share/media_data
The content under /mnt/media_share/media_data
is about a million of files of different sizes (it'd a CMS system). It's organized on directories e.g. /mnt/media_share/media_data/2000
, /mnt/media_share/media_data/2001
, etc.
I cant provide a log as I dont have it. But we have found that rclone compresses directories under this first directory level in a tar file, and then copies that to destination in S3 bucket.
I kind of understand if rclone compresses on the fly and places a tar in the S3 backend but I was kind of surprised to find my source compressed e.g. /mnt/media_share/media_data/2000/directoryA.tar
, /mnt/media_share/media_data/2000/directoryB.tar
, etc
Is this expected behavior when there is a huge number of files in the source? it's because the backend is S3?
Thank you,
Ruben