Hi,
My docker contain running Neo4J crashed due to no memory left on device with the following error: "Detected VM stop-the-world pause". I increased the memory of the VM and restarted the container. However the queries are running very slowly and timing out. I also added this to the config file:
server.memory.pagecache.size=175G
server.memory.heap.max_size=50G
server.memory.heap.initial_size=30G
Any suggestions to get it back up to speed would be really helpful.
Thanks!
What Neo4j version?
Memory configuration - Operations Manual suggests
It is recommended to set these two parameters to the same value to avoid unwanted full garbage collection pauses.
and yet you have configured with different values, for example
server.memory.heap.max_size=50G
server.memory.heap.initial_size=30G
also Tuning of the garbage collector - Operations Manual
The Compressed OOPs feature in the JVM allows object references to be compressed to use only 32 bits. The feature saves a lot of memory but is only available for heaps up to 32 GB. The maximum applicable size varies from platform and JVM version. The -XX:+UseCompressedOops option can be used to verify whether the system can use the Compressed OOPs feature. If it cannot, this will will be logged in the default process output stream.
and to which you have exceeded the suggested 32G.
Hello hello!
There's one query that's draining your heap. What kind of queries are u running?