SQL linked server returning error suddenly upon query

Question

We are facing error when trying to connect to and execute queries using a linked server. The SQL Server and the Linked Server are able to ping and telnet each other. Also testing the link returns positive. However, the following error is returned when we execute a query accessing the linked server:

OLE DB provider "SQLNCLI11" for linked server  returned message "Protocol error in TDS stream".
OLE DB provider "SQLNCLI11" for linked server  returned message "Communication link failure".
Msg -1, Level 16, State 1, Line 0
Session Provider: Physical connection is not usable [xFFFFFFFF]. 
OLE DB provider "SQLNCLI11" for linked server  returned message "Communication link failure".
Msg -1, Level 16, State 1, Line 0
Session Provider: Physical connection is not usable [xFFFFFFFF]. 
OLE DB provider "SQLNCLI11" for linked server  returned message "Communication link failure".
Msg 10054, Level 16, State 1, Line 0
TCP Provider: An existing connection was forcibly closed by the remote host.

I'm trying to prove that this is a network issue, but since the communication channel is open and ping and telnet work, the task fell back in my basket.

Any advice will be greatly appreciated.

Answer 1

This does look like a network issue, as per this question on MSDN the "Protocol error in TDS stream" indicates as such.

This seems to be a very similar question that was resolved by disabling TCP Chimney Offload (think link in that question directs to a veritas article that describes how to do it).

Which may be a quick (and possibly dirty) was of trying to resolve it. If that is not the problem, then it is almost certainly a network problem (maybe a nic problem).

--

Edit:

Might be worth making sure the native client is fully up to date (by updating SQL Server, which is no bad thing anyway).

Also, if you're still struggling, add some more information about your environment to your question e.g. OS version, SQL Server version & update level, information about your nics etc.

Answer 2

The issue was indeed network related and the network and data center team finally agreed. Right now, the root cause analysis is ongoing and the first culprit is a malfunctioning NIC on one of the VMs. There is also a DNS issue with one of the Always ON SQL Server Node.

However, we did a failover to another site, and the linked server is running again.