none
Multipoint 2012 Premium - NUServer64 messages and disconnects

    Question

  • Hi Folks,

    We are running Multipoint 2012 Premium in a virtual machine on a HP server (2012 Hyper-V)

    We experience a number of issues and Google never seems to know anything about the events that are logged.

    The clients are Atrust m320 zero clients and from time to time they randomly disconnect and reconnect with different behaviour.

    Pattern 1 - All devices disconnect at once, display "Waiting for assigned host" and then come back one after the other. Nothing is logged at all on the multipoint server

    Pattern 2 - Two devices go back to the login screen at the same time whilst they are in use. It can be any two devices and the following event sequences will be logged:

    "<00616B00_2> Device lock fail." Source NUServer64, event ID 4.
    "<00616B00_2> Connect fail." Source NUServer64, event ID 4.
    "<00616B00_2> Device does not exist." Source NUServer64, event ID 3.

    "<00616B00_4> VGA device not connected" Source NUServer64, event ID 3.
    "<00616B00_4> Device does not exist." Source NUServer64, event ID 3.
    "" Source NUServer64, event ID 3.

    There are numerous other NUServer64 messages logged all over the place, the system also complains about terminal service licensing. I am unsure if this is just noise or indicates that the Multipoint server is incorrectly trying to use TS Licensing - we certainly haven't told it to and there are no TS group policies

    Can anyone help me decipher the NUServer64 events? Is there a resource that documents them all for reference?

    Thanks,
    Colin


    Wednesday, November 27, 2013 12:27 PM

All replies

  • Colin,

    The TS licensing message can be ignored. The other errors though sound like you've got something going on with networking. When I've seen this (with one client and in my lab) it was because the zero clients were disconnecting and reconnecting. In one case this was because of unstable power situation. In another it was because the zero clients were being moved around from place to place. They are USB over Ethernet and USB doesn't like to be repeatedly plugged and unplugged. I've found that once the server is in this condition that they only way to get back to normal is to power it off and I mean all the way off as in unplugged from the wall, let it sit for a 10 count and then start back up making sure that all of your zero clients are connected first.

    Also many clients are we talking about here? They can be a maximum of 15 USB devices connected to your server at one time. That's zero clients + any other USB stuff. That's a Windows limitation.


    thanks, Amy

    Tuesday, December 03, 2013 7:58 PM
  • Thanks for the reply,

    It is unlikely to be networking unfortunately. We have a nice new gigabit smart switch that isn't reporting any errors and all server equipment is UPS protected.

    The other thing that makes me doubtful is when it affects two zero clients at the same time but not always the same devices. All the units are fixed and do not move and there are only 10. With it being a VM, no other USB devices are attached.

    In our case, the server itself never gets into a tizzy - there was one occasion where it lost communication with all zero clients and had to be restarted but that was before we replaced the switch. The devices always go off and come back on anywhere from a few seconds to a few minutes later.

    I posted the question in the partner forums as well and got told to ring PSS. I spent over an hour the other day being passed through various queues with differing levels of horrendous music and didn't get the support needed!

    Wednesday, December 04, 2013 9:23 AM
  • Given this additional information I would lean toward a driver or firmware issue. They've been coming out with updates to those pretty rapid fire for some of the zero clients.

    thanks, Amy

    Wednesday, December 04, 2013 2:32 PM
  • Thanks for the reply, alas there don't seem to be any updated drivers and the system has all other updates installed on it.

    The Atrust is based on the SMSC chipset - I can't find any newer than the ones released in February unless you know of any better links?

    Wednesday, December 04, 2013 2:40 PM
  • The ISVs are responsible for distributing the latest drivers. That said, there is sometimes a lag before they can release the latest updated drivers.

    For USB over Ethernet clients (both SMSC and MCT chipset based), Elite provides the USB over Ethernet portion of the driver.

    NuServer64 is the Elite component, so the problems you are seeing are part of that component.

    Elite has done numerous improvements here, and you can get the latest "generic" version of the Elite portion of the driver from the Connect site.

    https://connect.microsoft.com/wms/Downloads

    You will still need the SMSC or MCT based VGA driver, so see the notes on connect on how you might be able to upgrade just the Elite USB over Ethernet portion (basically start with the combined VGA/USB over Elite driver installed, then uninstall the old Elite driver and then update to the lastest Elite)

    Thursday, December 19, 2013 9:20 PM
    Moderator