How does it scale?

[JeffreyBenjaminBrown]

I've been using Semantic Synchrony, which uses Neo4j, to keep a knowledge graph of a few hundred thousand nodes. Could Dyna handle that much data? (The queries I want to make are pretty simple: "Show me all of X\s children and grandchildren," "Show me the intersection of the descendants of X and Y," etc.)

[nwf]

The current implementation scales very poorly, being a proof of concept. It is probably not well suited for anything but small test cases. @timvieira may have some actual numbers.

[j6k4m8]

[not a developer nor poweruser] Probably! I've used dyna for O(10K) rules and it didn't seem to choke or slow at all, though I don't have benchmark numbers. Running on 128GB RAM, so ymmv.

[timvieira]

As @j6k4m8 says, a couple-thousand edges should be fine. @JeffreyBenjaminBrown, If you pursue this, let us know how it goes!

[JeffreyBenjaminBrown]

Will do!