Duplicate ncalrpc RPC Protocols

[Hector Santos]

Our RPC server binds on ncalrpc and nacn_ip_tcp. On occasion, we get
reports from Window 2000 customers who report a 1740 (duplicate end point)
problem on the ncalrpc protocol.

We typically point them to KB articles and registry information setup to
make sure they have the protocol RPC client protocols.

With a new customer report today, I went a little further to find out why
some of our Windows 2000 customers are having this problem. I gave this
customer a simple utility just to list the available RPC protocols on his
W2K/Server machine. Well, it shows he has two ncalrpc protocols. This
explains why our RPC server has a duplicate binding error. It didn't expect
a second ncalrpc protocol.

Can anyone explain under what situation or circumstance why Windows 2000
will return duplicate ncalrpc protocols? I don't think this is normal or
possible. Or is it?

We checked the registry HKLM/Software/Microsoft/RPC/ClientProtocols and it
looks fine here.

Tips? Comments?

Thanks in advance

--
Hector Santos
Wildcat! Interactive Net Server
http://www.santronics.com

[Greg Kapoustin [MSFT]]

microsoft.public.win32.programmer.newtworks is the best place to ask RPC
questions

1740 RPC_S_DUPLICATE_ENDPOINT is usually returned when an RPC server is
trying to listen on an endpoint that is being listened on by someone else.

I am not sure what it means for the machine to have "two ncalrpc protocols".
Can you please explain in more detail what settings your tool retrieved to
determine that two ncalrpc protocols were registered?

Thanks,
Greg

--
This posting is provided "AS IS" with no warranties, and confers no rights.
Use of included script samples are subject to the terms specified at
http://www.microsoft.com/info/cpyright.htm

"Hector Santos" <nos...@nospam.com> wrote in message
news:ulUZ9Of$CHA....@TK2MSFTNGP11.phx.gbl...

[Hector Santos]

"Greg Kapoustin [MSFT]" <grig...@online.microsoft.com> wrote in message
news:3e937af9$1...@news.microsoft.com...

> 1740 RPC_S_DUPLICATE_ENDPOINT is usually returned when an RPC server is
> trying to listen on an endpoint that is being listened on by someone else.

right.

> I am not sure what it means for the machine to have "two ncalrpc
protocols".
> Can you please explain in more detail what settings your tool retrieved to
> determine that two ncalrpc protocols were registered?
>

RpcNetworkInqProtseqs() is used to obtain the available protocols. To
confirm the problem, I gave the customer a simple program that simply calls
this function and displays the vector of protocol sequences. It shows two
instances of ncalrpc.

This explains why our current RPC server is failing with the 1740 error when
it attempts to bind to the second instance of ncalrpc. Sure, a fix is
to not bind it again. But I want to find out why this occurrs in the first
place as the fix might hide the real problem or a more serious OS/networking
installation issue on the W2K machine.

With some of customers in the past, the problem also occured on Windows XP
when they have ZoneAlarm's TrueVector technology enabled. TrueVector
allowed the binding to occur the first time, but if the server was stopped
and restarted, TrueVector held on to the bind. In this case, it occurred
with ncacn_ip_tcp. The bug was reported to the ZoneAlarm people.

But this customer doesn't have ZoneAlarm, hence why the deeper analysis.

Thanks

[Hector Santos]

FYI, this turned out to be a customer's Dell rack server Windows 2000
installation issue. They had the server rebuilt and this cleared up the
duplicate NCALRPC protocols.

Thanks

---

"Greg Kapoustin [MSFT]" <grig...@online.microsoft.com> wrote in message
news:3e937af9$1...@news.microsoft.com...