What do you want?
We are evaluating HelixDB for object-storage-backed deployments and would like a documented way to run the local or self-hosted engine against an arbitrary S3 or S3-compatible object-store prefix.
Today we can run the public CLI locally with:
helix init local --path /tmp/helix-data --quiet --no-skills
helix start dev --port 8080 --disk --quiet
As far as we can tell, --disk starts a local Docker/MinIO-backed storage mode. We have not found a documented CLI or SDK path that lets us point HelixDB at a user-controlled remote object-store location, for example:
s3://my-bucket/my-helix-prefix/
Ideally this would support:
- AWS S3 bucket/prefix configuration
- S3-compatible endpoint URLs
- region configuration
- standard AWS credential-chain support
- access key / secret key / session token configuration where needed
- optional SSE / SSE-KMS settings
- restart/reopen behavior from an existing object-store prefix
A useful flow would be:
- Start HelixDB against a specific remote object prefix.
- Write data into that prefix.
- Stop the runtime.
- Restart HelixDB against the same prefix.
- Continue querying without importing or copying data back into a local-only store.
We checked the public CLI help for commands such as:
helix start --help
helix init local --help
helix add local --help
helix add cloud --help
and did not see a documented flag such as --storage-uri, --s3-uri, --bucket, --endpoint-url, or similar. helix add cloud appears to accept a gateway URL, but it was not clear whether that supports a user-controlled object-store prefix.
If this is already possible, could you point us to the exact command/API and documentation? If not, would you consider adding a documented local/self-hosted S3 configuration surface?
Feature Area
Infrastructure
Additional context
No response
What do you want?
We are evaluating HelixDB for object-storage-backed deployments and would like a documented way to run the local or self-hosted engine against an arbitrary S3 or S3-compatible object-store prefix.
Today we can run the public CLI locally with:
helix init local --path /tmp/helix-data --quiet --no-skills helix start dev --port 8080 --disk --quietAs far as we can tell,
--diskstarts a local Docker/MinIO-backed storage mode. We have not found a documented CLI or SDK path that lets us point HelixDB at a user-controlled remote object-store location, for example:Ideally this would support:
A useful flow would be:
We checked the public CLI help for commands such as:
and did not see a documented flag such as
--storage-uri,--s3-uri,--bucket,--endpoint-url, or similar.helix add cloudappears to accept a gateway URL, but it was not clear whether that supports a user-controlled object-store prefix.If this is already possible, could you point us to the exact command/API and documentation? If not, would you consider adding a documented local/self-hosted S3 configuration surface?
Feature Area
Infrastructure
Additional context
No response