do client need to replicate whole gbs of db ?

[yashwanth2804]

does replicating bulk dbs resources and time consuming, anyway can we limit upto few records
thanks

[RichardLitt]

Good question, @yashwanth2804.

Let's make sure to add an answer to the docs when someone provides one.

[mistakia]

I don't believe there is currently a way to limit replication. You may want to explore modeling your data differently (on orbit-db or even ipfs).

Generally speaking, when modeling data for OrbitDB you should strive to "partition" it where it makes sense - i.e. along access rights. As an example, in a twitter-like application, each user should have their own database instead of a global "tweets" database for everyone.

[yashwanth2804]

thanks for the reply,
I understood your explanation.say if i follow Justin X (random 🌟) on Twitter, where he makes tons of tweets a day.if i am gonna subscribe him after 2 yrs. As your mentioned pubsub model ,i subscribed for him so will i get megatons of tweets from his 1st twt.
i may limit in ui but how can i limit in loading those tweets.

please correct me if i am off the topic or missing something.

[mistakia]

If you have the tweets locally - you can limit how many entries are loaded by sending in an amount to db.load(amount):

const db = await orbitdb.feed("<justin's tweet OrbitDB address>")
await db.load(10)
// db now contains the 10 latest entries/tweets

If you don't have the data locally, I'm not sure there's a way to limit syncing to only the 10 latest entries.

[cristiano-belloni]

If you can't limit how many data you can replicate, it means that malicious peers could trick you to "follow" them, then just flood you with data. Which might be a huge problem in 99% of the use cases.

[aphelionz]

Moving to Field Manual to discuss further