Runtime Query Definition

Hi,

I am curious whether TigerGraph plans to develop a runtime query parser / distribution mechanism any time soon.

The advantages of being able to precompile and distribute parameterized execution plans for your queries are obvious, but what about use cases where the query plan cannot be known ahead of time? Using TigerGraph as a knowledge graph would functionally require the ability of users to define their own queries (in UI or whatever) to answer their questions, and users would be unlikely to want to wait around for their query variations to get installed on the cluster. Is there any plan to speed up query compilation / distribution time to the point it could be inline with execution?

Thanks!

Rob

Hi Robert,

Thanks for your question.

Now we are in the final stage of releasing our interpret mode if GSQL in TigerGraph 2.4. Interpret Mode allows the user runs the query without having to compile it or install it. It is quite handy when you write varieties of queries to explore the graph. Please stay tuned for our 2.4 release.

Thanks.

That’s very exciting! What is the general timeframe for the release of 2.4?

Our current goal is by the end of next week (06/21).

Robert,

You can take a preview of this interpret engine on this webinar.

Scripts are here

https://github.com/tigergraph/ecosys/tree/master/guru_scripts/pattern_match

Mingxi