NaraJeamfry/db_stress_comparison

I wanted to make sure what database to use for my current project. Because I feel like graph dbs will be faster in some cases that I know I need in the project, I want to also compare its performance in some cases.

DB stress comparison for Cosplaye.rs

I'm currently building a new application to serve as a social network to cosplayers. As such, users will be able to interact between them in multiple ways, like:

• Upload and manage pictures
• Tagging pictures with defined tags (that are searchable, and users can follow directly)
• Like, comment or otherwise interact with specific pictures
• Perform pretty complicated queries to all this data, for example:
  • Retrieve a list of pictures that share tags with a given picture, sorted by the count of tags a user follows
  • Retrieve a list of tags considered similar to the tags currently followed by the user, taking into account the list of pictures liked by the user and the tags of these pictures

These queries will probably be too much for a relational database, but theory says a graph DB will be really powerful in the sense of making these calculations fast enough. Additionally, databases like Dgraph provide ACID compliance and use minimal resources, so they should be safe to use for these use-cases.

So, I wanted to make sure what database I actually will use. To do so, I want to compare the performance of each DB paradigm in some example and edge cases.

How to build

TBD

Example schema (in GraphQL)

type User {
    username: String
    uuid: String! @id
    uploadedPictures: [Picture!]
    likedPictures: [Picture!]
}

type Tag {
    name: String
    uuid: String! @id
    relatedTags: [Tag!]
    taggedPictures: [Picture!]
}

type Picture {
    title: String
    uuid: String! @id
    isAbout: [Tag!] @hasInverse(field: "taggedPictures")
    uploader: User! @hasInverse(field: "uploadedPictures")
    likedBy: [User!] @hasInverse(field: "likedPictures")
}