To Cisco Nexus VSM VEM

Hello

When I try to go Cisco Nexus VSM the MEC connect get lost and never come back to the top again.

We have configured uplink port profile as mentioned in the Cisco Document "Install the Cisco Nexus 1000V software". Then, we added three profile more port for control and administration of the package VLAN. But as soon we will the NIC1 (control) of the connection to the lost GET VEM VSM and MEC never appears again.

-We missed to add the system of virtual local area network for the uplink port profile and the profiles of port for control and Packet of Cisco 1000V network.

Make a long story short - VSM can now be connected and also tied the MEC. But we constently receive messages "KERN-1-SYSTEM_MSG: a fall received executives of MSM in double - core" as soon as we add a second physical NETWORK adapter to the uplink port group.

Someone at - it a tip would be what is the problem?

THX

Tom

Tom,

I was able to recreate what you see. The problem is that just to add another NIC to the uplink without going through the uplink configured for a kind of switch teaming causes issues. Because the N1KV is essentially a switch she expected all NIC teaming to be a port channel or vPC - HM (virtual host port-channel mode). This must be set in the profile of uplink port. When you attach a second NETWORK card without going through the proper switch configuration, you can get the packets of streaming double and a host of other issues. This is why the MSM is complaining.

I don't know a lot about the 3010. If it only supports a port-channel vPC - HM is the way to go. vPC - HM will create a port channel virtual between the network cards on two different switches without having to make any changes on the physical switches. As long as the switches supported the CDP and the ports are configured the same it should work.

Louis

Tags: VMware

Similar Questions

  • Cisco Nexus 1000V Virtual Switch Module investment series in the Cisco Unified Computing System

    Hi all
    I read an article by Cisco entitled "Best practices in Deploying Cisco Nexus 1000V Switches Cisco UCS B and C Series series Cisco UCS Manager servers" http://www.cisco.com/en/US/prod/collateral/switches/ps9441/ps9902/white_paper_c11-558242.html

    A lot of excellent information, but the section that intrigues me, has to do with the implementation of module of the VSM in the UCS. The article lists 4 options in order of preference, but does not provide details or the reasons underlying the recommendations. The options are the following:

    ============================================================================================================================================================
    Option 1: VSM external to the Cisco Unified Computing System on the Cisco Nexus 1010

    In this scenario, the virtual environment management operations is accomplished in a method identical to existing environments not virtualized. With multiple instances on the Nexus 1010 VSM, multiple vCenter data centers can be supported.
    ============================================================================================================================================================

    Option 2: VSM outside the Cisco Unified Computing System on the Cisco Nexus 1000V series MEC

    This model allows to centralize the management of virtual infrastructure, and proved to be very stable...
    ============================================================================================================================================================

    Option 3: VSM Outside the Cisco Unified Computing System on the VMware vSwitch

    This model allows to isolate managed devices, and it migrates to the model of the device of the unit of Services virtual Cisco Nexus 1010. A possible concern here is the management and the operational model of the network between the MSM and VEM devices links.
    ============================================================================================================================================================

    Option 4: VSM Inside the Cisco Unified Computing System on the VMware vSwitch

    This model was also stable in test deployments. A possible concern here is the management and the operational model of the network links between the MSM and VEM devices and switching infrastructure have doubles in your Cisco Unified Computing System.
    ============================================================================================================================================================

    As a beginner for both 100V Nexus and UCS, I hope someone can help me understand the configuration of these options and equally important to provide a more detailed explanation of each of the options and the resoning behind preferences (pro advantages and disadvantages).

    Thank you
    Pradeep

    No, they are different products. vASA will be a virtual version of our ASA device.

    ASA is a complete recommended firewall.

  • Restoration of Cisco Nexus 1000V - Host-ID fingerprint

    Someone find some information about how to restore a Cisco Nexus 1000V?

    The license is the result of a fingerprint of the identifier of the VSM. In case we lose the VM with VSM or host ESX Server must be reinstalled, this print is different. So that would mean the licensekey need, it's be regenerated.

    Has anyone found information on it?

    Tom

    Q: can you a VSM manage it's own VEM?

    R: Yes

    ...

    Can Q: you a VSM vMotion?

    A: we do not recommend it.

  • Replacement of failing on Nexus 1000v VEM

    I was curious how others accomplish the replacement of a failed ESXi host who is an on the Nexus 1000v VEM. I did this procedure once and it seemed endless. The goal is to make the swap of the transparent to the Nexus 1000v (VEM even #, just different VMware UUID)

    -Transition to the standard vSwitches home

    -Remove the switch distributed via vCenter host (host of right click and delete)

    -Swap physically on the chassis

    -Find the UUID for new host (got esxcfg-info)

    -Install new MEC in the 1000v with this UUID

    -Host of replacement over 1000v that assumes the number of Vincent who was installed just to migrate

    Ben,

    Note - 1000v in relationship issues are better posted in the «Server network» forum  This forum is specific to the UCS.

    https://supportforums.Cisco.com/community/NetPro/data-center/server-network?view=discussions

    The procedure that you use is the right.  Another method is to remove the host of 1000v with elegance, then issue a 'no vem X"that removes the MEC does record from the MSM.  Exchange your hosts, then add back to the 1000v.  Is there a reason that you need the same UUID?

    Kind regards

    Robert

  • help required for cisco nexus 1000v

    Hello

    I have three esxi host in my environment and I want to integrate these hosts with cisco nexus 1000v switch.

    I installed vsm on host1 and adding the remaining host via vsm Update Manager. exchanges I have already create in SMV shown in the welcome that I've added to the vsm, but the port group is not shown on the host1 esx on which I have installed vsm, should I also add the host that contains MSM in the cisco nexus switch?

    I want to say that I have installed the MEC on any army three esxi. is it good?

    Hi Mohsin,

    Where did you read that? In the past, we have added the guests, including one who executes the VSM. Usually run us both VSMs (primamry and secondary) and add anti rules affinity so that the two VSMs are on different hosts. I'm not a person CISCO, but having worked with CISCO engineers, we had no problem with what you have just mentioned. It would really be a waste of host in my opinion. I don't see why this could be a problem... As long as you have all your trade (PGs for your packages VSM etc etc) in place, you should be able to add all hosts in my experience.

    Follow me @ Cloud - Buddy.com

  • 40g of Cisco Nexus-FET

    I'm trying to choose the right cable to connect to a Cisco Nexus 40 G - FET (on a Nexus 6004). I saw on Cisco's Web site that says the FET transceiver has 12 fibres inside. Yet, when he broke out to connect to four FETs of 10G on a FEX (2232TM-E), only 8 fibers would be used.

    so... What type of connector DFO buy? A 12 FTU fiber or fiber 8 FTU? Are they interchangeable? I might as well use an 8 or a connector DFO 12 fiber on my fan-out cable?

    Cisco uses 8 fiber lit even if standard DFO is of 12 fibers. Please refer to Figure 23 here.

    As described further down in this document (figures 25 and 26) some (cited Corning) providers offer only fiber 8 DFO and which can be used for interconnection as well.

  • Switch Cisco Nexus 6004 removes more than 300 bytes IP packets

    Hi all

    We have a circuit of wave level 3 10 G running between two switches Cisco Nexus 6004. The circuit came online between our two data centers (in the same city) without problem.

    When attempting to ping to the remote-end 10G interface, it works very well with packets of 64 bytes. CDP is enabled and that we see the CDP information remote switch. However, if we increase the size of the ping to more than 300 bytes packets, we lose 1 in every 20.

    We settings MTU verifed, type of cable of 10G and duplex settings.

    Level 3A tested clean and we will move forward with more testing.

    Any ideas on the problem? We feel the carrier out and end to test with their testers. But so far, the circuit's own test. I was not sure if it is something related to Cisco. I am at a loss at the moment.

    Thank you.

    Mike

    Hi Mike,.

    There is nothing wrong with your switches or circuit.  NEXUS devices have a default COPP on the control plan that limit the size and the amount of traffic that must be the CPU process.

    http://www.Cisco.com/en/us/docs/switches/Datacenter/SW/6_x/NX-OS/security/configuration/guide/b_Cisco_Nexus_7000_NX-OS_Security_Configuration_Guide__Release_6.x_chapter_011001.html

    HTH

  • Cisco Nexus 3548/3524 SNMP Interface traffic problem

    Hi all!

    I would like to share with you the problem I've encountered SNMP interface respect tracking on the Cisco Nexus 3500 of traffic devices.

    We is transferring our network infrastructure in the Nexus and we encountered a problem with the interface bitrate of tracking traffic on Nexus 3548: deltas SNMP counter ifHCInOctets and ifHCOutOctets show no valid data.

    We have investigated this issue and figured he could be linked with the refresh rate of the meter.
    I measured, by snmpwalk, the time between simple increments the counter ifHCInOctets to Interface VLAN and it apears to about 140seconds. But for physical interfaces (Eth), the interval of interface SNMP counters are 15 seconds and with sampling frequency of the 1960s for the queries we can still observe the problems on graphs (see attached screnshot).

    We get the same results when testing the question on various physical devices (3548 Nexus) on versions of different software - including the last being.

    And for Ethernet interfaces (same NMS - ZABBIX-config):

    We tested a wide Huawei s6700 switches and the range of Cisco IOS devices and there is no such problems.

    The main problem is that we cannot monitor our crucial nexus, via SNMP devices (as you can see from the attached chart).  Results are the same when configure us Cacti or Zabbix to monitor Nexus devices.

    The results of this monitoring not having meritorical value or use.

    We are working on this case with Cisco TAC, and they manage to reproduce the problem in their LAB environment. But for now, they cannot provide any solution and they claim that this isn't a bug. If we get any final statement for this case, I'll post here for your information.

    I'll be more than happy if you can post ideas that help me solve the problem described.

    Thank you in advance!

    Michal

    Looks like a bug to me.  His return of erroneous data.

    Maybe try asking your file be traced back to the team leader.

  • Cisco Nexus to use authentication Radius AAA using Microsoft 2008 NPS

    I have a Nexus 7010 running

    I was wondering if you can help me with something. I'm having a problem with the approval of the order through our aaa config. We have not an authentication problem of command approval that does not work. From what I've seen and read Nexus NX - OS 6.x has not all orders for the aaa authorization, unless you configure GANYMEDE +. My basic config is below if you can help would be much appreciated.

    > ip source interface mgmt radius 0

    > key RADIUS-server XXXXX

    > host X.X.X.X key radius server authentication XXXXX accountant

    > RADIUS-server host X.X.X.X XXXXX key authentication accountant aaa

    > authentication login default group aaa authentication Radius_Group

    > RADIUS server logon group console local aaa Radius_Group

    > server X.X.X.X

    > server X.X.X.X

    > mgmt0 interface-source

    Also nobody how to configure Microsoft 2008 NPS as Raduis server to work with Nexus? I read a few post that suggests to change the

    Shell: roles = "vdc-admin" in the value field of the attribute in the RADIUS server

    Anyone know if it works?

    Thank you

    I haven't used NPS before but sounds like you are on the right track. As Ed mentioned in his post, GBA, you can set the type of protocols that you will accept during an authentication session. Authentication Nexus sessions is considered as PAP/ASCII, so you should be good to go. I don't have a Nexus switch to test with, but if you can use wireshark to capture the session and see the exact protocol / method used. However, I am sure that PAP is the way to go:

    http://www.Cisco.com/c/en/us/TD/docs/switches/Datacenter/SW/4_1/NX-OS/se...

    I also found the link that you might find useful:

    http://www.802101.com/2013/08/Cisco-Nexus-and-AAA-authentication.html

    Thank you for evaluating useful messages!

  • Cisco Nexus 5548UP switch current Version 5.2 (1) N1 (4)

    Hi, we are currently have Cisco Nexus 5548UP switch current Version 5.2 (1) N1 (4)

    It is not listed as affected version:

    Known affected releases:
    (5)
    5.2 (1) N1(8A)
    6.0 (2) N2 (5)
    7.0 (3) N1 (0.125)
    7.0 (4) N1 (1)
    7.1 (0) N1 (0.349)

    The fixed versions known as follows:

    Known fixed releases:
    (7)
    6.0 (2) N2 (4.3)
    6.0 (2) N2(5A)
    7.0 (1) ZN (0.615)
    7.0 (5) N1 (0.173)
    7.0 (5) N1 (1)
    7.1 (0) N1 (0.371)
    7.1 (0) N1 (1)

    It needs to be updated to the latest versions?

    If Yes, what version of the fixed known better adapt our current version

    Cisco Nexus 5548UP switch current Version 5.2 (1) N1 (4)

    Thank you.

    I had the same question, so I opened a folder of TAC to find the version of the software. I was told that the software listed as available in the bug report are versions of NX - custom OS and that there was no version patched for the 5548p. The support engineer also said that the updates of Nexus 5 k should not until November or December. The best advice I could get was to subscribe to the bug report and wait to be notified of an available upgrade.

  • Cisco Nexus 5596 does not start after NX-OS upgraded from 5.1.3.N1.1 to 6.0.2.N2.5 [CSCun66310]

    Hello friends,

    I want just to save your valuable time & effort, while Nexus 5596 upgrades so I wanted to share some useful information that you can have a quick glance before your Cisco Nexus 5596 device updates.

    Recently, I ran into trouble when upgarding 5596 Nexus data center switch does not boot and after investigation of the Cisco TAC it comes out to be Bug described below.

    https://Tools.Cisco.com/bugsearch/bug/CSCun66310/?reffering_site=dumpcr

    My way of upgrading the device was: 5.1 (3) (1) to 6.0 (2) N2 N1 (5).

    As this upgarde involves an upgrade of Bios and power purposes, I had some doubts to get this properly upgraded, but he had more twist involved in this upgrade I got to know after contacting Cisco TAC for my death 5596 Nexus device and got info if you have version impacted in your Nexus 5596 device and you plan to upgrade , then you migh RUN in this serious BUG aftre which device will not come at all and only Option left is to replace with device of RMA.

    Material affected PID versions Versions updated

    UCS-FI-6296UP                                            1.0                                       1.1

    N5K-C5596UP                                              1.0                                        1.1

    N5K-C5596T                                                 1.1                                         1.2

    You can check with hardware version command from your device below.

    5596 # display sprom sup | H/W Inc.

    H/W Version: 1.1

    So please be careful when you plan your Nexus 5596 update and check over things that this Bug is not present anywhere in the docuemntations upgrade unfortunately.

    Hope this will help and save valuable time of a person.

    This bug affected me. Before upgrade you the equipment, to track:

    It's procedure, there will be no allocation to apply.
    Is attached to the case of the dplugin for each NXOS, check out the name associated with each version.
     
     
    Step 1
     
    -------------------------------
    Copy the file dplugin and update, load Plugin
    Copy the ucd - update.tar to the bootflash
    Copy the debug plugin into the bootflash
    Load the image via the load control
     
     
    Note You can access the command line without leaving the plugin to debug via command vsh
     
    Example:
     
    n5K-1 # dir bootflash:. grep debug
    3023275 23 Jan 15:28:48 2013 ucd - update.tar
    283423 23 Jan 15:12:28 2013 nuova-or-dplug - mzg.5.2.1.N1.5.bin
    n5K-1 # load bootflash:nuova - or - dplug - mzg.5.2.1.N1.5.bin
    Loading of plugin version 5.2(3a)
    ###############################################################
    WARNING: debug plugin is for engineering for internal use only!
    For reasons of security, plugin image has been removed.
    ###############################################################
    Debugging properly load-plugin!
    Linux (Debug) #.
     
     
     
    Step 2
     
    -------------------------------
    After you load the dplugin file into NX - OS, please follow the steps to update the module of TI.
     
     
    Copy the ucd - update.tar of bootflash to the tmp directory and unzip the file
     
    Linux (Debug) # cp /bootflash/ucd-update.tar/tmp
     
    Linux (Debug) # cd /tmp/
     
    Linux (Debug) # tar - xvf ucd - update.tar
     
    UCD.bin
     
    Address_91_A.csv
     
    Address_92_A.csv
     
    Linux (Debug) #.
     
    Step 2
     
    ----------------------------------
     
    Run the ucd.bin file:
     
    Linux (Debug) #./ucd.bin
     
     
     
    You will see the prompt on the version 1.1 update.
     
    --------------------------------------------------
     
    principal: correctly upgraded TI inst 2 to version 5.
     
    New SPROM HW version 1.1. Charging must reflect in 'show the sprom' output.
     
    Release Linux (Debug) #.

    The problem is how to get the plugin. Only Cisco can provide this.

  • Cisco Nexus switches

    I'm looking to deploy a series 5100 Cisco NEXUS switch at 10 Gbps.

    I know that the Nexus is supposed to work with the converged network adapter (for 10 Gbps FCoE, etc.), but can it operate without an ANC?

    I want to put some passthrough 10 Gbps modules in my Dell m1000 chassis and the cables directly to the Nexus switch.

    I know that the Nexus is perhaps overstated for this solution, but it is a step in the UCS solution for us.

    Thoughts?

    James

    Hi, you don't need special drivers for "low latency" 10 Gbit ethernet on a 5 k.

    for example, to switch non-nexus 5 k

    PING 10.10.10.1 (10.10.10.1) 56 (84) bytes of data.

    64 bytes of 10.10.10.1: icmp_seq = 1 ttl = 255 time = 0,530 ms

    64 bytes of 10.10.10.1: icmp_seq = 2 ttl = 255 time = 0.618 ms

    and a nexus 5000 with a qlogic 8152

    PING 172.16.78.3 (172.16.78.3) 56 (84) bytes of data.

    64 bytes from 172.16.78.3: icmp_seq = 1 ttl = 128 time = 0.150 ms

    64 bytes from 172.16.78.3: icmp_seq = 2 ttl = 128 time = 0,134 ms

    Oracle rac cluster will fly!

  • Why is - that someone would need Cisco Nexus 1000v when DvSwitch is so Kool

    Why is - that someone would need Cisco Nexus 1000v when DvSwitch is so Kool

    Or is it something that DvSwitch cannot always do that Cisco Nexus 1KV possible?

    Use of 1kV Nexus are clear enough, if you want to segregation, the advanced settings of COS, use Cisco VSG, etc. etc., you must use Nexus 1kV. But if you do not use one of these, why would you pay more money to use a Nexus 1kV, while you can use dvSwitch, giving you more or less the same basic features. After all, the 1kV has been developed using the dvSwitch framework.

  • Update Virtual Center 5.0 to 5.1 (using Cisco Nexus 1000V)

    Need advice on upgrading production please.

    current environment

    Race of Virtual Center 5.0 as a virtual machine to connect to oracle VM DB

    3 groups

    1: 8 blades of ESXI 5.0 IBM cluster, CLuster 2: 5 IBM 3850 x 5

    2 cisco Nexus 1000v of which cluster only 1 use.

    I know that the procedure of upgrading to 5.1

    1. create DB SSO, SSO of installation

    2 upgrading VC to 5.1

    3. install WEB CLient set up AD authentication

    IT IS:

    I have problems with the Nexus 1000? I hope the upgrade will treat them as he would a distributed switch and I should have no problem.

    He wj, treat the Nexus as a dVS.

  • Configuration of Cisco Nexus

    Hi, wondered about the Configuration of the Cisco Nexus 100v.

    First of all, the MSM will have three vNIC used to:

    1. the Control Interface - NIC1

    2. Management Interface - NIC2

    3 packet Interface - NIC3

    The management interface is on the same VLAN as VMconsole (service console) because it is used to communicate with vCenter and ESX hosts for management traffic?

    Addresses IP should be assigned for control and packet Interfaces?

    Thank you

    When you use L3 control we encapsulate the control traffic and the package on the default management interface. Only IP you need is the IP address that you have assinged to the management interface 0. Interfaces and control packets never get used and never need a IP address.

    Louis

Maybe you are looking for

  • What does the acronym FROSTED for?

    Hello I see the acronym FROSTED or OP so everywhere used instead during the sale to load of the cables for the devices, but I can not find any of the terms online dictionary which defines it.  Someone can tell me what he means and means? Thank you Wo

  • My iPhone to send international sms without inform me please solve it

    Dear Apple support,Please, I lost my money in my iphone, cause my iphone to send international sms without informing me. Please would solve the problem A.S.A.P. Best regards Nabil Al Badani TZ جهاز الـ الخاص بي iPhone

  • On the Satellite A100-646 DVD drive replacement

    Hi all I want to replace the DVD drive on my laptop. (I never solved a lot of problems)Do you know if there is a list of compatibility for the A100-646 model? Model to suggest me? Thank you very much. ConcerningFabrizio

  • First installation on Satellite A500 - 14 k does not end

    Hello world I'm French, Taylor my name is, I 26.I just got my new PC, a Satellite A500 - 14 k. The first facility started more than 2 hours ago and I still have a blank screen in which it is written: "Please wait... The end of the installation may ta

  • HP Officejet 6500 a Plus - E710n, loses the wireless connection

    Mac OS X 10.7.3 installed with a usb cable When the printer is on, it is available for a few seconds (I am able to print/scan only once), then it becomes unavailable even if the printer is not available for the Mac, it is connected to my wireless net