Recently, we tried to switch our OrientDB 3.0.18 standalone installation to a very small distributed cluster with 2 nodes (following the distributed configuration as suggested by OrientDB’s). Setup was ok; inserted Data into one node; data was synced; everything seemed to be ok. But, when we had to run first updates on our data, we got an error “Quorum not reached” for some of the data objects. However, the objects which could not be updated were rather big (many fields and some of the fields contain big, complex json strings). Other (small) data objects were updated through the cluster without any problems.
To temporarily solve the problem, we put the “writeQuorum” value from “majority” to “1” on all nodes (asynchronous writes). No error occures any more. But, any idea what could be the problem with synchronous writes?