-
Notifications
You must be signed in to change notification settings - Fork 19
support for S3 object store #5
Comments
Hi @aleks-mariusz , Please let me know if I can be of any assistant. |
plus one for this, I'm working on it |
I'm considering using rclone to do the whole backup job, rclone support tons of cloud storage, we it can save a lot of developing time |
@TAvardit what do you consider a drastic image size change? Doing a test of just installing aws cli into the existing image it looks like it goes from:
largely..won't be able to say get that down a ton from there, due to python dependencies needing to be installed for the cli utilities. @LeoQuote rclone is interesting, but requires some looks like large reconfiguration for anyone across the board, adding the aws cli utilities grows it a decent bit, but would support the two main clouds. |
There is a fork with RClone: https://github.com/dbccompany/docker-nexus-backup |
It seems to me that this project only supports google's object store..
is there a technical reason for this, or does the author simply not have access to/interest of an S3-compatible object store to test/implement this (i.e, something like Minio would work, I'd imagine).
I'm considering rolling out this set of nexus-related containers at our org's kubernetes cluster, but the lack of s3 support is a bit of a non-starter for us as we're fairly invested in S3 as a cloud storage/backup target.
So rather than simply complain, i could look into implementing this myuself. However first, would a PR to support S3 even be something of interest to the org/maintainers of this project?
I'd imagine it would just involve something like a.) install boto (to provide the aws s3 utility), b.) update the entrypoint script to use that utility instead of gsutil depending on the bucket URI prefix.. and c.) test test test the heck out of it manually before submitting said PR
Please let me know however if there's no interest in having this functionailty (such as for support reasons), so as to save me the time of doing this work (as i can perhaps just create our own fork within our org if the travelaudience devs don't want to own this going forward).
The text was updated successfully, but these errors were encountered: