Should we use global traversal source in gremlin script directly?

Amiya <amiyakr...@...>


Is it recommended to use global traversal source to span traversal in gremlin script directly?

Or should we create a new transactional source for each gremlin script submission?

If there is any error during script execution through global traversal source for a fetch query, does it rollback automatically similar to bytecode connect? Or should I have to always wrap script in a try catch block?

Do we have a known scenario, when JG server memory does not clear even after a script evaluation (read op) complete?

I am using JG 0.3.1