I'm just getting started with OrientDB and trying to determine if & how
OrientDB scales for my use case, so any suggestions would be greatly
appreciated.
Here's some numbers to frame what I'd be trying to model in the graph.
- I will have approximately 35 billion vertices.
- Each vertex will have 4 properties (2 longs, an int, and 1 string).
- I will only need 2 edge types (a 1:1 and a 1:many relationship).
- I actually don't need a single graph for all 35 billion nodes, I could
get away
with 3000 separate graphs (approx 11 million nodes per graph).
since my application would always be querying 1 of the 3000 areas of
the graph.
Sorry I can't divulge more on what I'm modeling.
As a test, I generated a small graph (1 of my 3000) and ended up with
with 2 million vertices and 2 million edges which was 1.1GB on disk.
As a ballpark figure I estimated my complete data footprint to be approx.
20TB.
Any ideas, general suggestions on architecting such a (large?) distributed
graph?
Alan
--
---
You received this message because you are subscribed to the Google Groups
"OrientDB" group.
To unsubscribe from this group and stop receiving emails from it, send an email
to [email protected].
For more options, visit https://groups.google.com/groups/opt_out.
