You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Nex should be able to start workloads based on layers described by an OCI manifest. Additionally, we should make it easy for developers to bundle up the various OS- and CPU- specific versions of a given tag (where applicable).
As such, the following is a list of proposed enhancements:
When a workload start request has an artifact URL that uses the oci:// scheme, the appropriate CPU- and OS-specific binary (layer) should be downloaded from the registry
Figure out the best place to put registry credentials for accessing secure registries. This probably makes sense in the node configuration ?
nex art should be a subcommand that helps people push binaries into an artifact in a registry
add a binary to an artifact
remove a binary from an artifact
push/pull binaries
Use case
Nex should be able to accommodate users that still want or need to integrate with OCI registries (this does not cover running OCI/docker workloads... only the artifact management)
Contribution
No response
The text was updated successfully, but these errors were encountered:
Note that this does not supercede #395 - this applies to oci:// workload URLs. Having the OS and CPU suffixes on plain keys in a NATS object store will still apply to nats:// artifact URLs.
Proposed change
Nex should be able to start workloads based on layers described by an OCI manifest. Additionally, we should make it easy for developers to bundle up the various OS- and CPU- specific versions of a given tag (where applicable).
As such, the following is a list of proposed enhancements:
oci://
scheme, the appropriate CPU- and OS-specific binary (layer) should be downloaded from the registrynex art
should be a subcommand that helps people push binaries into an artifact in a registryUse case
Nex should be able to accommodate users that still want or need to integrate with OCI registries (this does not cover running OCI/docker workloads... only the artifact management)
Contribution
No response
The text was updated successfully, but these errors were encountered: