LMS 4.0.1 user tracking acquisition problem

[fbender]

Hi all,

I have a problem with user tracking as I see IP phones only on a few of my switches.

Credential Check runs without any issues, DNS is working for the devices (both name -> IP and IP -> name)

I have deleted and readded the devices and run a discovery afterwards.

What I found is: going to acquisition (Admin> Collection  Settings> User Tracking> Acquisition  Action) and do a Device Selection I can select the switches but when I then start the acquisition I get an error "unreachable Device".

Deivce Unreachable Report shows no unreachable devices.

Do I miss something?

Thanks

Frank

[Joe Clarke]

This error means ANIServer was unable to identify any layer 2 VLAN ports on the device in question.  This can happen if the device in question is not a layer 2 switch or if all of the ports are in a layer 3 interface mode.  This can also happen if Data Collection did not complete.  You should run a new full Data Collection then see if you can acquire the end hosts.

If you have any etherchannel ports in your network that are terminated on Cat2K, Cat3K, or Cat5K switches, you will encounter CSCto06189.  A patch is available at http://www.cisco.com/cisco/software/release.html?mdfid=283434800&flowid=19062&softwareid=280775103&os=Windows&release=4.0.1&relind=AVAILABLE&rellifecycle=&reltype=latest to fix this.

[fbender]

Thanks for the hint!

I have an issue with the ANI sever, when I try to connect from my browser it stops at about 30% percent and then timeouts somewhen - wanted to look into it later...

I did a full data collection but it did not change anything - and I get the data from some devices but not all. All devices in questions (working and not working) are 4507R connected by L3 Etherchannels to the same pair of 6513s.

I also get the info from all 3550s, 3560s and 3750s at remot sites.

Regards

Frank

[Joe Clarke]

The Connecting to ANISerber hang is most likely due to a firewall between your client and the server.  TCP ports 42342 and 43242 must be open between the client and server.  You can test this by telneting to those ports from the client.  If the connections hang then timeout, you will need to talk to your network security team to get those ports opened up.

You might still try applying the Data Collection patch I mentioned.  Any etherchannel terminated on an affected switch will cause Data Collection to crash, thus not properly processing all the devices.

[fbender]

No fw - I can telnet to the ports - my client is Win7 64bitJave 1.6.0_24 while the server is Win2003 64bit with Java 1.6.0_22.

When I run the Topology Services in a RDP session from the server it is working and showing the boxes as reachable.

I also installed the patch run the datacollection again but still fails with device unreachable for the acquisition.

[Joe Clarke]

You must use a 32-bit version of the Java Runtime.  It's best to install the JRE that comes with LMS (from NMSROOT/htdocs/plugin).  This version includes the necessary CORBA files.  You also need to make sure the short hostname of the LMS server is resolvable on the client.  That is, the following command executed on the client should give you a blinking cursor at the top of the DOS terminal:

telnet HOSTNAME 42342

Where HOSTNAME is the short hostname of the server.

How do these unreachable switches appear on the Topology Map?

[fbender]

I am not sure what you mean by Topology Map - sorry I am a newbie

I have attached n-hop view, topology services view L2, device center view

all are looking normla to me

[Joe Clarke]

Start a packet capture session using Device Center and the Packet Capture tool.  Filter on all traffic to this device.  Then try to perform device-level acquisition.  Post the resulting sniffer trace file.

[fbender]

sorry cannot find a packet capture tool - but if I want to start acquisition I get an unreachable will it start anyway?

[Joe Clarke]

The packet capture tool is in Device Center under the star icon above the device summary.  There should be some querying of the device.

[fbender]

here is a capture from the NAM in between - does this help - will look into the cws cap as well

[Joe Clarke]

This doesn't appear to be showing any of the packets that would be sent after starting a device acquisition.  There should be at least one query for sysObjectID, and that is not here.  You need to run the trace for a few seconds after the acquisition is started.

[fbender]

might it be it is not startting at all because of the status "unreachable"

I installed winpcap and restarted the serber but it seems I cannot start packet capture - no reaction on hidding "OK" - I have attached the window I see

[fbender]

this nam capture was running for about 10min

[Joe Clarke]

Post the vlanData.xml and portsData.xml files from NMSROOT/campus/etc/cwsi.