Infinispan cache inconsistency for massive concurrent operations

Question

I'm struggling with Infinispan cache consistency during a massive update of value on one key.

My cache is defined like this (I tried other options and combinations but without success):

<replicated-cache name="refCounterCache" mode="SYNC" >
    <locking isolation="READ_COMMITTED" write-skew="false"/>
    <versioning scheme="SIMPLE"/>
</replicated-cache>

The test case is very simple. On cache defined like this:

Cache<String, Integer> refernceCounterCache = cacheManager.getCache("refCounterCache");

In one of nodes I set initial key value for key as 1000;

I call on every node code that 2000 times increment value for key:

IntStream.range(0, 2000).forEach(i -> {
            try {
                refernceCounterCache.merge(key, 1, (vOld, vNew) -> vNew + vOld);
            }
            catch (Exception ex) {
                log.error("Unexpected error during value merge!");
            }
        });

When I run the above lambda code on a single node cluster (one program instance) everything works as expected (value 3000). The same is for two nodes in cluster - I get the same expected result: 5000

Problem starts when I have more than two nodes. In case of 3 nodes incremented concurently I get unexpected results e.g. 6973 instead of 7000.

Anyone know where the problem is? Maybe wrong type of cache? Maybe wrong locking options?