Hey, Im considering

Nebula Graph

for a new

OpenTelemetry

big data project. And I wonder if

Nebula graph

can handle it. the main

Vertex

will be called

Span

and we expect to add 6K spans per second. Each

Span

has an average of 25 attributes. some are intrinsic and will be part of the

Vertex

. The other can change, so I thought It will be best to create an

Attribute Vertex

and link each

Span

to its related attributes. That mean around 24K

Edges

per second. Assuming 1 month of data gathering, we can expect

16 bil Span

and

400 bil Attributes

There are other `Vertex`/`Edges` but they wont reach such numbers. There will be max of 20 simultaneous querying users. Query should be fast enough and measured in few seconds. Can

Nebula Graph

handle the `Write`/`Storage`/`Read` of such scenario?

It’s great to hear that you are interested in learning about and using Nebula Graph in projects with large amounts of data. Regarding your question, the answer is yes. In terms of data volume, Nebula Graph supports a scale of 16 billion spans and 400 billion attributes. However, this is based on the condition that your server configuration and quantity increase along with data expansion. You also mentioned query speed and efficiency issues, which may require understanding the enterprise version of Nebula Graph that has been optimized for queries to improve efficiency. As for reading and writing, we have many peripheral tools for data importation available on our official website.

Thanks Xiaolang for your answer. I will contact Nebula team for further details