1

I'm using the Gremlin language to query JanusGraph (and other database systems) and I'm typically sending my queries in the ByteCode format using the traversal op processor. However, for some queries, I need transactions (aka sessions) to batch multiple reads/writes, and for that I should be using the session op processor.

The problem - the traversal processor accepts queries in gremlin bytecode, whereas the session op processor accepts queries as gremlin strings. Is there any way in which I can make queries which are both transactional (as I need a sequence of multiples reads and writes), and are sent as ByteCode?

I'm mainly asking since I discovered that queries running as ByteCode via the traversal processor, have significantly lower overhead time than the same queries being ran as strings with other processors (about 30ms difference).

Thanks in advance!

1
  • stephen mallette, i think there exist some solution even through not perfect currently. please refer: issues.apache.org/jira/browse/TINKERPOP-2252 and i also think session is essential because many user need this ability,and noe4j and many other systems also support session Commented Jul 2, 2019 at 3:35

1 Answer 1

1

In TinkerPop 3.5.0 it seems they did add bytecode transaction support: https://tinkerpop.apache.org/docs/current/reference/#transactions

Sign up to request clarification or add additional context in comments.

Comments

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Start asking to get answers

Find the answer to your question by asking.

Ask question

Explore related questions

See similar questions with these tags.