Date
1 - 3 of 3
org.janusgraph.diskstorage.PermanentBackendException: Read 1 locks with our rid but mismatched timestamps
Ronnie
Hi Mark,
Thanks for the suggestions. Narrowed down the issue to JanusGraph's support for Azul Prime JDK 8, which generates nanosecond precision for Instant.now() API, as compared to the millisecond precision for OpenJDK 8. So the issue was resolved by applying the patch for https://github.com/JanusGraph/janusgraph/issues/1979. Thanks! Ronnie |
|
hadoopmarc@...
Hi Ronnie
No idea what is going on here, but just being pragmatic: the bin/janusgraph.sh script starts Cassandra, Elasticsearch and Gremlin Server using the conf/janusgraph-cql-es.conf graph configuration. As you can check in the janusgraph distribution files, this config does not specify the graph.timestamps property (so defaults to MICRO). So some things to try:
|
|
Ronnie
Hi,
Environment - JanusGraph 0.5.3 on JDK: 1.8 - Backend: Cassandra 3.11.3 running on JDK 1.8 Warning and error during first time server startup
Graph configuration:
Note: Explicitly set graph.timestamps=MICRO ; when setting this to NANO as suggested here https://stackoverflow.com/questions/58916854/janusgraph-janusgraphexception-could-not-commit-transaction-due-to-exception-dur i get the following error:
Any pointers why this time resolution mismatch is happening? Thanks, Ronnie |
|