Skip to content

Make DGraph distributed #14

Closed
Closed
@manishrjain

Description

@manishrjain

Both loader and query processor would need to support distribution.

Loader
Depending on number of servers, loader would only store edges corresponding to fingerprint(attribute) % N on each server.

Query Processor
Query processor would use the same formula to find the right server holding the attribute required, and appropriate send the query portion to that server, and consolidate results.

Metadata

Metadata

Assignees

No one assigned

    Labels

    kind/featureSomething completely new we should consider.

    Type

    No type

    Projects

    No projects

    Milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions