For our project we need to insert huge amounts of data into Janusgraph and do some traversals for our analysis. The traversal part works fine. But we have been having some problems with fast inserts and/or parallel inserts. We keep on getting the following exception:
Caused by: org.janusgraph.diskstorage.locking.PermanentLockingException: Expected value mismatch for KeyColumn [k=0x 16-165-160-118-108- 30- 78- 78-201, c=0x 0]: expected=[] vs actual=[ 48-141] (store=graphindex)
We always do a read + insert/update.
Any ideas what is going wrong here?
Has anyone used Janusgraph with Flink?
Is Janusgraph suited to do streaming inserts (read + update/insert) into it?
