com.orientechnologies.orient.server.distributed.task.ODistributedOperationException: Request ... didn’t reach the quorum of n, responses: [node: node_2 success,node: node_3 success..... The error is generral : To avoid: Do not update the same record in another process to avoid conflicts and lock the record The same process with many threads you need to lock the command or use OrientGraphFactory or ThreadLocal to update the data -When the cluster dies (locked down), you must stop all writes and rebuild the cluster (copy the database or use auto-sync, copying the database file is a snap for many machines). master) -If a node (server) crashes (hardware …) you do not restart while writing, just stop, do not start. Stop all writes and then start all servers in the cluster, but step by node master....1,2, slave 1,2.. If you have big data, you must increase resources (RAM / CPU) to avoid deadlock when one processor is too slow and not respond to another node. Above is the very helpless button that you have to do with distribution