Upgrade from prespock to spock with 2 buckets in prespock cluster fails

Steps to reproduce: With Java SDK 2.5.0

1. Create a 4.5.1 cluster with 1kv, 1kv/index/query
2. Create 2 buckets
3. Run kv operation against 1 bucket
4. Keep #3 and add a Spock kv node

Expected:

No error

Actual : Throughput gradually down to 0 with continues TimeoutException till rebalance is completed. TimeoutException continues after rebalance is completed. And here is the exception thrown from get/set operation after adding the spock kv node.

at com.couchbase.client.java.util.Blocking.blockForSingle(Blocking.java:77)
at com.couchbase.client.java.CouchbaseBucket.get(CouchbaseBucket.java:148)
at com.couchbase.client.java.CouchbaseBucket.get(CouchbaseBucket.java:143)
at com.couchbase.sdkd.cbclient.GetCommandContext.doOneCommand(GetCommandContext.java:76)
at com.couchbase.sdkd.cbclient.CommandContext.execIter(CommandContext.java:308)
at com.couchbase.sdkd.cbclient.CommandContext.execute(CommandContext.java:353)
at com.couchbase.sdkd.server.SdkServer.executeCommand(SdkServer.java:178)
at com.couchbase.sdkd.server.SdkServer.handleRequest(SdkServer.java:199)
at com.couchbase.sdkd.server.SdkServer.run(SdkServer.java:254)
Caused by: java.util.concurrent.TimeoutException

*Note1: this does not happen with prespock to prespock.

*Note2: libcouchbase and .NET works fine with prespock to spock upgrade case with 2 buckets

Related to Go SDK(GOCBC-235), I see 0x08 couchbase.status from the spock node being added when I run with Go SDK but there was no couchbase packet returned from spock node when I run with Java.