I have been trying to understand how the caching of content from other Lemmy instances works. From what I have gathered, the local Lemmy instance will automatically download and store posts made to any communities that are followed by users on the local instance.

To me, this seems somewhat unsustainable in the long term - I am aware of the fact that it’s only storing the text of the posts, and not any media. I’m curious if it’s possible to configure the local instance to only cache the stored data for a certain amount of time (it might be better to just periodically purge the entire cache with a cronjob, or something); however, the data that I would like to store permanently is posts to any other community by users on the local instance, as well as posts made to communities on the instance (I have a suspicion that the communites data is permanently stored by default).

  • ClassyHatter@lemmy.world
    link
    fedilink
    English
    arrow-up
    16
    arrow-down
    1
    ·
    1 year ago

    As far as I know Lemmy stores only text locally and images and such will be linked to the external instance. Text doesn’t use much disk space, so that shouldn’t be a big a problem. Sometimes when you browse Lemmy, you notice posts that have broken links to images. It’s because the other instance is down, but you can still see the text portion of the posts on your home instance.

    • PriorProject@lemmy.world
      link
      fedilink
      English
      arrow-up
      3
      ·
      1 year ago

      The full firehose of the lemmyverse will grow your DB by about 2GB per day, plus an additional 8GB per day in thumbnails. And the rate is growing frantically. It’s not necessarily trivial, depending on how much you subscribe to.

  • emhl@feddit.de
    link
    fedilink
    English
    arrow-up
    6
    ·
    1 year ago

    AFAIK the cached images are only of thumbnail size and get deleted after a few months

  • Oliver Lowe@lemmy.sdf.org
    link
    fedilink
    English
    arrow-up
    4
    ·
    1 year ago

    May be worth having a look at the ActivityPub protocol. It’s the way Lemmy instances (and other stuff!) communicate with one another. From there I think it will be clearer how a single Lenny instance could behave. https://activitypub.rocks

  • lemmy@lemmy.kodemystic.dev
    link
    fedilink
    English
    arrow-up
    2
    ·
    1 year ago

    I have just setup an instance of Lemmy on my server. So if I federate to a community, all its posts will be stored on my server, also images and so on?

    • PriorProject@lemmy.world
      link
      fedilink
      English
      arrow-up
      4
      ·
      1 year ago

      The terminology is that you federate with an instance and subscribe to a community.

      And yes, if you sub a community, your instance will mirror posts and comments to your local DB and re-serve them to the public unauthenticated internet. On a single-user instance, anyone can browse the communities on your server and infer what your account subscribes to, including whatever porn, piracy, and other legally questionable shenanigans based on what gets replicated to your server.

      • lemmy@lemmy.kodemystic.dev
        link
        fedilink
        English
        arrow-up
        1
        ·
        1 year ago

        Thank you. What about comments and posts that get updated/changed? Will my instance also reflect those changes?

        • PriorProject@lemmy.world
          link
          fedilink
          English
          arrow-up
          1
          ·
          1 year ago

          Yeah, they get federated as a new edit message. Not sure whether they get stored in the db as edit entries, or if the DB actually just modifies the original… but edits federate.