I’m evaluationg if dGraph can be the main database for my project since the query languaje it’s very expressive and perfect for graphQL. So far everything looks good but I could not find any information regarding big nodes.
Does big nodes (don’t confuse with nodes with many edges) degrade query performance? like nodes with predicates containing 200kb of text or more?
Nodes in Dgraph are technically an abstraction from a set of predicates/Subjects (mean a set of Posting Lists). Which are the smallest data in Dgraph. They are distributed in groups and recorded via BadgerDB in Key Value.
So, each value has limit size of 2GB (check this topic Storing blobs in Dgraph). And the performance is guaranteed by BadgerDB. So 200kb is very light for Badger.
I don’t think so. Query process can’t be affected by the value size.
In isolation, Value sizes can not cause performance loss. Only a critical set of factors could.
Cheers.
To execute the query, Dgraph does not send the data to the client. So, this statement is a bit misleading.