DCNM Discovery

Hi
I'm having issues with DCNM discovering some Nexus 7K's. Shallow discovery seems fine. I enter the Nexus and its snmpv3 credentials in
DCNM> Admin>Data Sources>LAN
I enter the device as a single device in a switch list and it finds it as manageable. I select the device and it goes into Discovering.
Sometimes it completes this part and allows me to go to Deep discovery. Other times I get snmp timeout messages. I have had it complete a deep discovery and show up on my topology map only for it to disappear a day later again with snmp timeout errors. I've confirmed the snmpv3 credentials work with a 3rd party MIB browser which lets me test credentials and all is ok with them.
My first question is
The DCNM server seems to have an incorrect time. Is this a possible issue ? If so how do I set the time ? I've looked all over the GUI and can't find any place to set it. Is it taken from the VM platform or is it set independently ?
Are there any other issues that may cause discovery problems ?
Thanks, Stuart.

Further troubleshooting has led me to put a capture on the Nexus to see if that sheds any light on what's going on. I get the following output from the capture.
172.31.1.1 -> 172.30.1.1 SNMP Source port: 59596 Destination port: snmp[Malformed Packet]
That's from the DCNM server to the Nexus.
Stuart

Similar Messages

DCNM Discovery Issues

Hello Community,
Can someone please help determine why I'm unable to discover our NX7's and NX5's, using DCNM-LAN. From the screen shot when I click on 'Click here to discover the devices in DCNM', I'm taken to the web page showing the devices that have been discovered (see screen shot 2). However, the devices don't appear in the Cisco Prime DCNM-LAN.
And I imagine the reason I'm not able to see any devices in the topology view is because I'm not able to discover any devices?
Please help.
Cheers
Carlton

Hi Carlton,
After discovery you need to perform an "deep discovery" for the devices.
You can find this here:
admin>data sources>LAN
Mark the task and press the "edit LAN' pencil icon.
Then select the device and OK. After this the devices will show up in DCNM - LAN.
Cheers
Alex

DCNM Discovery Seed address

When I use DCNM to discover a Nexus 5548, entering its mgmt0 address (172.16.x.190) as the seed, DCNM finds it but lists it as 1.1.1.190. It takes a long time to discover for some reason, too. When I open the fabric, it cannot access the switch until I change its address manually, at which point it seems to fully access the device. However, the fabric seed address is still listed as 1.1.1.190.
Any idea what's going on here?

I have this problem too.
I'm adding the seed address as the vlan management address, however, after an initial dialog it then always tries to talk to the management interface - which i've set up as private vPC keepalive link (1.1.1.2 in my case).
I guess its picking this up from the FCS database, but I can't find anywhere how to change it;
nexus# sh fcs database
FCS Local Database in VSAN: 1
Switch WWN : 20:01:01:01:01:01:01:01
Switch Domain Id : 0x67(103)
Switch Mgmt-Addresses : snmp://1.1.1.2/eth-ip
http://1.1.1.2/eth-ip
Fabric-Name : 20:01:01:01:01:01:01:01
Switch Logical-Name : nexus
Switch Information List : [Cisco Systems, Inc.*N5K-C5596UP*5.2(1)N1(7)*20:00:00:00:00:00:00:00]
Switch Ports:
Any takers?

Dcnm - 6500 snmpv3 discovery timeout

Hi DC Experts,
I'm trying to fullfill a new install for our new DC with DCNM 6.2 (win2k8 64-bit). N5K, N1K are easily imported via the DCNM discovery
with SMNP v3 auth. Now i've run into problems with the 6500/sup2t.
SNMP v2 gives no problems but i'm not able to achieve this with v3. From the windows server cmd it's possible to do a snmpwalk with the
snmpv3 username/pass.
Does anybody have an clue wat the problem can be?
Cheers,
Alex

Further troubleshooting has led me to put a capture on the Nexus to see if that sheds any light on what's going on. I get the following output from the capture.
172.31.1.1 -> 172.30.1.1 SNMP Source port: 59596 Destination port: snmp[Malformed Packet]
That's from the DCNM server to the Nexus.
Stuart

FM/DCNM Slow Discovery

For the past few months I have had an issue with slow discovery of my MDS9134's via Fabric Manager (SA). We have a pair of 9124's at another site that discover and open instantly, however the pair of 9134's, which are local, take about 5-10 minutes each.
I have recently upgraded the switches and FM to 5.0.7.....but this did not help the issue. I also tried installing and using DCNM 5.2.2a but it has the same problem.
I have tried uninstalling FM/DCNM completely and reinstalling.....and when I try do to my initial discovery, I get this messsage: "Fabric is not discovered yet. Keep waiting?" If I select YES to continue waiting, I receive that message a couple more times and then the fabric is finally discovered and I can open it and everything seems to operate normally. However, once I close FM and reopen it, I have the same delay just trying to open one of the already discovered 9134 fabrics and I receive this message: "Fabric_xxxxxxx is in the process of being discovered. Please wait before trying again."
Does anyone have an idea on where to start my troubleshooting for this issue?
****We recently connected our MDS9134's to our UCS 6248's via a FC Port Channel.....I wonder if this may be part of the issue??***

Hi David,
By default, FM/DCNM uses SNMPV3. In the MDS, this can be seen as the "snmp-server user" ID and password.
By default, the UCS does not have SNMP enabled nor will it allow you to use "admin" as the SNMPV3 userid.
You can only have one userid and password on the fabric tab on the control panel for FM/DCNM, so the same one has to be defined and used for all switches ( including the fabric interconnect on the UCS ) in the fabric.
It looks like you have defined the userid and passwords:
(1)However, these have role network-operator and you are gonna want this to be network-admin, so that you can do zoning ( if you are using the default role definitions ).
(2)Make sure this password is being used on both the fabric and open tabs on the Control Panel for FM/DCNM ( there are two places ).
http://www.cisco.com/en/US/docs/unified_computing/ucs/sw/gui/config/guide/1.4/UCSM_GUI_Configuration_Guide_1_4_chapter6.html#task_49D35C84025E4495BBE50BE01CA7D0D0
Not sure what version of FM/DCNM or UCS Firmware you are using but check these bugs:
CSCtx45591 UCS discovery fails via DCNM
CSCtq73035 DM does not support UCS 6248 Fabric Interconnect starting from UCS 2.0 release.
Best regards,
Jim

DCNM-LAN 6.2(3) Discovery tasks failed

Hi,
After trying since several days i did not come to a solution. I add for discovery in DCNM-LAN a Task with 2 defined Switches. Now i am facing the problem that after starting the task it stucks and the Task has no more Switches in the list and i cant start deep discovery or rediscovery failed with Device 0 error. I have a Oracle 11.2 on a Windows 2008 R2 Server and for DCNM i use the same Servertype as for the Oracle DB. Users on the Switches are authenticated by ACS 4.2 via Tacacs. I took a trace with wireshark on the DCNM Server and and what i can see is that the discovery process on the Switch take a log time until the discovery starts. Also DCNM Server trys to connect with http to the switch, which fails (no feature http-Server) and i also can see the SNMP Requests.
Any help will be appreciated
Michael
Sent from Cisco Technical Support iPad App

you can first debug on your server to see if your snmpv3 is working. i installed this snmp toolkit on a red hat linux server that is also running the DCNM services: http://www.net-snmp.org/download.html
then i used snmpwalk to test if snmpv3 is working properly:
snmpwalk -v 3 -a md5 -l authNoPriv -u snmpuser -A Snmp255! 192.168.100.32    // the IP is of the switch to manage.
when verified that snmpv3 is working then try it in DCNM.
notice that snmpv3 requires a complex password with: lower case letters, upper case letters, numbers and a special character.
The user "snmpuser" must be able to login to the switch via SSH to privilege-exec mode.
but the commands to set up snmpv3 on the switch are following:
snmp-server host 10.113.113.99 traps version 3 auth snmpuser          // the IP is of the DCNM server
username snmpuser password Snmp255! role network-admin               // here you locally define the snmpv3 user as network admin, because the same user is also used for SSH access.
In your case I see you haven't enabled snmpv3, but 2c instead.

DCNM 5.2 - Device discovery question

DCNM 5.2(1) - Nexus 7000 with only one VDC configuration. Would it be possible to discover the equipment through a vlan interface (in-band management) instead of using an ethernet interface?

Hi Nilton,
Yes DCNM supports discovery through SVI as long as the device is reachable through SSH.
Thanks,
Ravikumar

DCNM v5.2(2),S104 Discovery for 6500s Very Slow

It's been discovering a 6513 for over 90 minutes. Is this an expected duration or is something likely malfunctioning or is it just the sheer # of interfaces that it must process? The Nexus 5k's that it has discovered progressed fairly quickly.

As far as I know, this is not normal.
Suggest to take a look in the DCNM-LAN server log file :
C:\Program Files (x86)\Cisco Systems\dcm\jboss-4.2.2.GA\server\dcnm\log\server.log
If no extra information in the log file, I would open a TAC case.
Paul

Prime DCNM 6.1(2) - LAN client not working

Hello Everyone,
I've recently installed a Prime DCNM 6.1(2) to a Windows 2008 R2 SP1 64 bit version. It seems to be working fine, apart from the DCNM-LAN client.
If I login with the Java client, I got the GUI, however it remains blank - I can see the menu, but no data at all. It seems it cannot cummunicate with the DCNM server. DCNM-SAN client, the WEB client and the Device manager works fine, but I just cannot make this work.
Tested it from Win2k3 SP2, Windows 7, and the server itself, with Java 1.6.31, 1.6.37, 1.7.13.
Tried to install DCNM on our LAB as well, but I've experienced the same.
Any hints appreciated!
Kind regards,
Gabor

Hi All,
      I also have met this issue! However, I want to know whether the deep discovery must need SNMPV3 or not. My client just gave me the N7K username and password, I used the information to discovery the N7K successfully and the dcnm-lan client still can't display anything.
     DCNM version is 6.1(2) and window platform is win 2008 R2 standard 64 bit
    Thx in advance!
     After setting the SNMPv3, we still find nothing in the dcnm-lan client as given below:
Regards,
Tira
消息编辑者为：tira li

DCNM 6.1(2) AD Integration

Hi All
I am trying to integrate DCNM 6.1(2) into Microsoft Active Directory.
I am trying to find some infomation on how to set up the Admin Group Maps to Map AD Group membership to the DCNM Role.
Have looked through the following docs but nothing on how to set this up.
DCNM_Fundamentals_Guide.pdf
b_Cisco_DCNM_Security_Configuration_Guide_Release_6.x.pdf
b_Cisco_DCNM_Troubleshooting_Guide.pdf
Anyone got any material . experience in this?
Appreciate the other option is to use an the "Attribute" option with AV-Pairs, but I'd prefer using the "Admin Group Map" option.
Thanks in advance.
Colin

Hi All,
      I also have met this issue! However, I want to know whether the deep discovery must need SNMPV3 or not. My client just gave me the N7K username and password, I used the information to discovery the N7K successfully and the dcnm-lan client still can't display anything.
     DCNM version is 6.1(2) and window platform is win 2008 R2 standard 64 bit
    Thx in advance!
     After setting the SNMPv3, we still find nothing in the dcnm-lan client as given below:
Regards,
Tira
消息编辑者为：tira li

Monitoring nexus 5K using DCNM (linux platform)

Hi guys,
since some time i installed on a linux server. (the server is shared with Netapp onCommand / DFM).
nothing to say about installation and discovery of the nexus switches.
they i read the manual to understand how to collect performance data... and here start the problem ( in my opinion the documentation is not very clear and miss the basics).
i did the following steps:
1. created a flow on 1 nexus switch for 2 zones traffic direction both. (i tried to create for all zones all cards but i've got a out of memory error and im not able to understand if its a server problem of a nexus problem)
2. created collections on DCMN-> admin.>collections for both nexus
3. started performance manager
the result after some time (ciouple of hours) is:
on DCNM->performance->end devices->all ports
i see all NaNB (usually means the snmp response is blank or -1)

what happens is when the server boots the HBA sees the array and boot lun, but when you start the ESX 5.1 install the server initiator logs out of the SAN and the boot lun disappears. If you use ESX 4.1 everything works fine. I opened a TAC case and they confirmed that the fnic driver in ESX 5.1 will not work with the 4700's directly connected to the FI.

SNMP Traps and DCNM 6.1(2)

Hello Everyone,
I have DCNM 6.1(2) installed on a Windows Server 2008 R2 to monitor and manage 4 Nexus 7018 (12VDC's).
I currently have one problem left. All snmp traps send from the Nexus Switches are all dropped and none processed, and i currently rely on the syslog messages, but would prefer the snmp traps.
I start wondering if the snmp traps are only for SAN devices?
Or If I am doing something totally wrong. Port for the traps is 2162. Discovery via snmpv3 works... Traps are sent and recieved but dropped.
Perhaps someone could shed some light into this.
Thanks
Pol

Hi,
Yes, DCNM for SAN supports forwarding of alerts using Cisco Call Home, EMC Call Home, email, and Simple Network Management Protocol (SNMP) traps.
DCNM LAN supprts Syslog and E-Mail.

Queries about DCNM

I have configured my DCNM install using the SVI interface. I have changed the option to managed from unmanaged, but when I login It seems like I am not seeing all the info. First what I notice is that against the switch name it says timeout. Secondly I do not see any ports neither Eth nor FC. I see the VSANs but no ports associated with this VSAN? I would like to know if I am doing something wrong. Please help.

Hi Everyone,
i know that this problem is very old but maybe this will provide a solution:
The DCNM uses per default the mgmt IP instead of using the SVI. Best Practice Scenarios will tell you to use the mgmt0 Interface for vPC Keepalive, because this is a free Copper Port and you dont have to use an expensive Nexus Port. Anyway DCNM will prefer this mgmt0 port for polling in SAN Discovery.
To solve your problem just go into the Server Options of the DCNM Admin -> Server Properties. Scroll down to the section Polling:
I hope that this could solve your problem.
Regards,
Sascha

DCNM-SAN supporting FI

Using DCNM-SAN with UCS (2.1.2a); I created on all FC devices (MDS and N5k) a user "dcnmuser" with netadmin rights. I also created a snmp v3 user "dcnmuser" with UCSM. Then I did a fabric discovery with DCNM-SAN; I see the fabric, MDS, N5k and FI, however, the FI have a red arrow, which means not manageable.
Q. what could be wrong ? or is this really supported ?

Understood, thanks Robert
We only want to monitor UCS-FI with DCNM-SAN.
The release notes 6.1.2
http://www.cisco.com/en/US/docs/switches/datacenter/sw/6_x/dcnm/release/notes/dcnm_6_1_relnotes.html
are quite misleading:
Introduction
Cisco Data Center Network Manager combines the management of Ethernet and storage networks into a single dashboard to help network and storage administrators manage and troubleshoot health and performance across the following product families that run Cisco NX-OS software:
•Cisco MDS 9500 Series Multilayer Directors and Cisco MDS 9200 and 9100 Series Multilayer Switches
•Cisco Nexus 1000V Series Switches and Cisco Nexus 1010 Virtual Services Appliance
•Cisco Nexus 2000 Series Fabric Extenders
•Cisco Nexus 3000, 4000, 5000, and 7000 Series Switches
•Cisco Catalyst 6500 Series Switches
•Cisco UCS 6100 and 6200 Series Fabric Interconnects
•Cisco FireWall Service Module (FWSM)
Cisco DCNM-SAN Release 6.1(1b) and later releases do not support Cisco Nexus 5000 Series switches and UCS devices running Cisco NX-OS Release 4.1(3) or an earlier release.
However, Table 9 Hardware Features Supported by Cisco DCNM Release 6.1
doesn't show UCS FI !

DCNM v7.1 fabric already exist error message

Hi my friends:
Do you konw where storage san device in dcnm because when i try to add another this show me that the same device exist,but i can found in screen,
I attached the error message.
Thank,
Best regard.
Marco

Hi Marco,
DCNM will discover the entire SAN fabric from a single discovery job. By entering the IP address of an initial seed switch, DCNM finds and adds adjacent switches. It is likely that the .34 switch was found by one of the 2 previous discovery jobs. There are two places in DCNM to determine if the switch has already been discovered.
You can find a list of all devices managed by DCNM, including SAN devices, in the Web GUI under [Inventory > Switches]. Additionally, you can launch the DCNM-SAN java client to view the fabric as well as configure the devices. To launch the SAN client, open the DCNM Web GUI and find the 'DCNM-SAN' button in the upper right corner of the screen. The SAN client accepts the same user credentials as the Web GUI. Once logged into the SAN client, select the discovered fabric and 'open' it to view all of its devices.
-Eric

DCNM Discovery

Similar Messages

Maybe you are looking for