I am setting up Orient DB master-master cluster environment and facing below two issues:
1)Maximum connection reached
2)Concurrent modification exception while writing record.
I am using Orient Graph factory with max 20 connection for my 4 client apps. Also all the graph related call is surrounded by try and finally having graph.shutdown(), so where connection is getting leaked?
Another issue which I am observing is writing a vertex to class in clustered environment. It hits OConcurrentModificationException and once I try to print the IDs I saw the record got started with #60:0 and while commit during graph.shutdown() it has id as #60:1. I do not have any concurrent thread running. So is DB sync up can modify the record id.
Same write record once i tried using single node it worked fine.