DeliciousBluewhale87 , Hi!

I think you can have models/artifacts automatically copied to a location if the experiment is initialized with
output_uri
For example:
task = Task.init('examples', 'model test', output_uri=' ')What version of ClearML are you using? I'd suggest upgrading to the latest 🙂

we can always use the latest Clearml.
We were thinking of a use case for a client, who has Sonatype Nexus in their environment ? Could we leverage on it, or would we need minio instead ?

In our local setup, we use minio though ?

DeliciousBluewhale87 , what would be your use case? You can save various artifacts/models/files to nexus via the previously mentioned method (You can also configure this in your clearml.conf file under sdk.development.default_output_uri="" )

CostlyOstrich36 :
They mentioned that they already have a Nexus backend. So just was wondering if we could use it for storage purposes.

DeliciousBluewhale87 , I believe so, yes 🙂

Thanks. 👍

DeliciousBluewhale87 basically any solution that is compliant with S3 protocol will work. An example:
output_uri=" :PORT/bucket/folder"Are you sure Nexus supports this protocol ?
I "think" nexus sits on top of a storage solution (like am object storage), meaning we can use the same storage solution Nexus is using.
Just to clarify we do not support the artifactory protocol Nexus provides for storing models/artifacts. But we do support it as a source for python packages used by the agent.

Quick update Nexus supports direct http upload, which means that as CostlyOstrich36 mentioned, just pointing to the Nexus http upload endpoint would work:
output_uri="http://<nexus>:<port>/repository/something/"See docs:
https://support.sonatype.com/hc/en-us/articles/115006744008-How-can-I-programmatically-upload-files-into-Nexus-3-