Also check the ESXi NIC port connected to the switch port has hybrid mode, and pass all relevant VLAN.
↧
Re: Can not Ping Secondary Ip in VM Created By ESXI 5.1
↧
how the packet flow will work from VM in one Host to VM in other Host
Hey Guys
Could you please help me to understand how the packet flow will work from VM in one Host to VM in other Host
I wanted to understand more about what will happen when virtual switch will see the traffic and pass it to physical switch
I understand packet will follow as below
Example :VM1 > vSwtich > ESX1NIC > PSwtich > ESX2NIC > vSwtich > VM2
Both VM are in same Subnets and a single physical switch is connected to both ESXi's
VM1 packet will contain
VM1 Source MAC
VM1 Source IP
VM2 Destination MAC
VM2 Source IP
The process of packet flow ( as per my understanding, please correct if wrong):
1. As the vswtich does not find entry in its MAC cache, it forwards the traffic to pSwtich via ESX1NIC
2. pSwtich will have MAC entry of the destination VM2 which will be mapped to the port of the ESX2NIC
3. pswtich will forward the traffic to vSwtich of ESX2 and vSwtich will lookup its MAC Cache and forward the packet to destination VM2
Questions :
1. will Physical Switch have MAC entry of multiple VM's residing on vSwtich mapped to a single physical port?
Thank-you in advance
Packet Flow from VM in one Host to VM in other Host
Message was edited by: SK21090
↧
↧
Backup Virtual NIC stop responding after backup on Linux VM
Hi All,
My Name is Mithilesh Singh, I am working in a corporate organization and having ESXi 5.5 virtual infra.
There is some issue reported for Linux VM and team is saying that backup NIC on VM automatically stop responding.
Issue reported -:
1. There is backup scheduled which execute on a particular time(example: 1 AM). Once backup initiate second time it get failed because Backup team report issue that client (Linux VM) stop responding on network level and backup master unable to connect it.
2. Unix Team disable\enable the backup NIC and it start to ping(this will happen every next scheduled).
3. We verified VMs and found that some of the VMs having old network adapter and someone with latest vmxnet3 and tools are updated.
4. Issue has been reported in FLEX Chassis and standard ESXi.
5. We checked the Virtual Switches(VDS) and everything are ok..
6. we have ESXI 5.5 and All host are having latest patch installed.
7. We have mix environment Windows and Linux but facing issue only on Linux VM. This issue is reoccurring only on some VMs not all.
Could you please help in it.
↧
Legacy session is deprecated and might not be supported in future releases.
When I am trying to add port mirroring in my distributed switches, I see below error. vCenter version is 6.0.0, 3018523. Version of distributed switch is 6.0.0. I have similar vcenter and dvSwitch with same version in our DR where I can add port mirroring without any problem. Can someone help?
↧
ESXI/Vmware network setup, just getting started.
I got 3 hosts, each have Cat5 port and a 10G fiber port. I am new to VMware but have worked with Hyper-V for the last 5+ years. We are moving to VMware and I am trying to get an eval system up with vsan, vsphere, vmotion and eventually VDI. I can manage all hosts through CAT5 port which is connected to cisco 2960. Just trying to set up the net backbone on the fiber ports. All host see fiber port as connected. They all connect to a cisco 4500x. I want to separate the traffic into 6 vlans (VM, Vsan, Vmotion, mgmt., Repl, VDI) and I created these vlans on the cisco side. Just need help getting it setup on the VMware side. Maybe I am thinking about it all wrong but I thought these could all run over the fiber port in a trunked config. Please help guide me in the right direction
Also the fiber card is a quad port so eventually I will create a etherchannel on this.
ESXI/VMware 6.5.
↧
↧
Re: ESXI/Vmware network setup, just getting started.
Since you're using your CAT5 and fiber ports, I will recommend you create a vSwitch using the CAT5 ports and use that vSwitch for management purpose, and then create a second vSwitch with your fiber ports and use that vSwitch for the remaining traffics. And avoid create Etherchannel, that will just add complexity to your environment and not much benefits, just use LBT teaming, see the following blog post for additional details about LBT x Etherchannel: Etherchannel and IP Hash or Load Based Teaming? | Long White Virtual Clouds
↧
Re: ESXI/Vmware network setup, just getting started.
OK great, but how do I trunk the port on the VMware side or do I need a separate physical port for each VLAN?
Say I got a vSan VLAN. I have assigned VMkernal port on host 1 of 192.168.0.2 on vswitch1
I got a VM Vlan. I have assigned a VMkernal port on host 1 of 192.168.1.2 on vswitch1
On host 2 I do the same thing except the IP's are 192.168.0.3 and 192.168.1.3
Both hosts connect to a switch with 1 10g fiber port. Both vSwitch1's have this 10g physical fiber uplink port.
I created a VLAN int on the cisco switch for each vlan/network with ip of .1
ON cisco side I trunked the 10g port.
Is this the correct way to set it up or do I need a separate physical port for each VLAN?
↧
Re: ESXI/Vmware network setup, just getting started.
The switch port on physical switch should be configured as trunk port to allow traffic from different VLANs, this way a single physical port can handle traffic from more than one VLAN. At VMware side, for each VLAN you will need to create a Port Group and assign the specific VLAN for each Port Group.
VMware works different from Hyper-V, in Hyper-V you assign the VLAN on network virtual adapter properties of the virtual machine, but on VMware you assign the VLAN to a Port Group, and once you connect your virtual machine to that Port Group, the virtual machine will use the VLAN assigned to the Port Group.
↧
Re: how the packet flow will work from VM in one Host to VM in other Host
1. will Physical Switch have MAC entry of multiple VM's residing on vSwtich mapped to a single physical port?
Yes, on physical switch MAC address table you will see a lot of entries pointing to the physical switch interface where do you have your ESXi host connected. Where each MAC entry can be a MAC address of a virtual machine or ESXi VMkernel interface.
↧
↧
VM stops pinging after it moves to Distributed switch portgroup.
I create a distributed switch and move both hots to the DPortGroup. But when I migrate the XPVm to the DPortGroup then this vm stops pinging. But it networks shows it is connected. When I changed it’s adapter to VMnetwork then it starts pinging but the moment I changed it to DPortGroup then it stops. Where is the problem occurred which I am unable to find. Please guide. I have attached the screenshots.
↧
Re: VM stops pinging after it moves to Distributed switch portgroup.
Can you confirm that the uplinks used for the dvSwitch are configured correctly, and that the port group to which you migrate the VM has the correct settings (e.g. VLAN-ID)?
André
↧
Re: VM stops pinging after it moves to Distributed switch portgroup.
Thanks for the point you have given me. I was totally confused of why my VM is not pinging to Host Esxi servers and Domain. I went through a webpage docs where somebody suggest an extra LAN Card need to be added in both Esxi Servers to build up Distributed switch. I did that and then problem started. You have told me to check the port group settings and dvSwitch settings. My total setup then was in mishmash condition just because of that extra NIC which I added and made it an extra VMkernel. Then I follow your guide line, made a complete new setup after complete uninstallation of domain and others. I Build up the systems with new ip addresses and domain name. After that everything starts working fine including vMotion & VsphereHA is also working fine. I just want to know from you that is it necessary to put an extra LAN card in both Esxi servers to build up a distributed switch?? Please guide me again.
↧
Re: VM stops pinging after it moves to Distributed switch portgroup.
There's basically no requirement for additional network adapters. You can use the wizard to migrate from VSS to DVS.
... just because of that extra NIC which I added and made it an extra VMkernel.
Why did you add a VMkernel port group? VMkernel port groups are only required for host networking, i.e. not for VM networking.
André
↧
↧
Re: VM stops pinging after it moves to Distributed switch portgroup.
Dear Sir
In my scenario there are Two Esxi servers I have installed in VMware Workstation. Both Esxi servers are communicating between each others via VMkernel port. I have only one LAN port in both server so I thought if that NIC used in VMkernel port then by which port the distributed switch will be communicating. That thinking forced me to add an extra LAN card to the Esxi server. Now I understand that NIC was the main culprit to spoil everything. Thanks to you heartily for advising me. Actually I was trying to solve it since last Friday. There were many sleepless nights also. Do you have any LinkedIn account?? Could you please add me??
↧
vDS LAG + another active uplink for the same distributed port group
I am using vSphere 6.5 with Enterprise Plus license.
Here is a simplified version of my vDS topology:
- There are 2 hosts: H1 and H2
- Host H1 has 2 NICs, and host H2 has 4 NICs
- The first NIC on each host, i.e. vmnic0 is dedicated for management and vMotion
- Other NICs are for VM traffic
- There are 2 distributed port groups in vDS: Management and VMs.
- There are 2 uplinks in vDS: uplink1 and uplink2
- Teaming property for Management port group shows uplink1 as active and uplink2 as "not used"
- Teaming property for VMs port group shows uplink2 as active and uplink1 as "not used"
- Physical NICs dedicated for management, i.e. vmnic0 on both hosts (see item 3 above) belong to uplink1
- Physical NIC dedicated for VM traffic on host H1 belongs to uplink2
- The 3 physical NICs (i.e. vmnic1, vmnic2, vmnic3) on host H2 are combined into LAG which is visible beside uplink1 and uplink2
My goal is to keep only one VM-related distributed port group VMs and let it use uplink2 for VMs running on host H1 and LAG uplink for VMs running on host H2.
I thought that I need to make both uplink2 and LAG being active uplinks for VMs port group... but it doesn't work (web client doesn't allow it). As a temporal resolution I had to map VMs to uplink2 (via setting teaming properties) and create another distributed port group, say VMs3 and map it to LAG uplink also by means of teaming properties.
Any help and/or clarification on how I can keep a single distributed port group for all VMs is appreciated. Thank you.
--
Simon
↧
can't have a connectivity using a dvportgroup
hello,
i have a vm connected to a csr1000v with a trunk(vlan 200-208) but can't get a ping between the two vm, strangelly i have the same thing 2 vm with the same thing and these works fine , do you have some ideas?
thanks a lot
↧
vsphere standard switch nic teaming
Hi all,
Nice to meet everyone here.
I am new to the vsphere/esxi virtualization and is currently using VMware ESXi, 6.0.0, 3380124
My host has a standard vswitch created & uplinked to two physical NICs. The two physical NICs on the host are then in turn connected to a pair of stack physical cisco switches.
I am having some confusion about NIC teaming on standard vswitch and I hope gurus here can guide me along..
q1) does using NIC teaming on standard vswitch requires the compulsory / equivalent setup of static etherchannel (port-channel x on) on the physical switch ?
or rather
can NIC teaming on standard vswitch works withoutany etherchannel setup (be it static or lacp doesn't matter) on the physical switch ?
-- reason for asking this is I saw in VMware KB that only "Route based on IP hash" works for static etherchannel. but how about other load balancing method ?
q3) just for a final confirmation, LACP only works for VDS even in my current ESXI version right ?
Hope to hear from you guys soon.
Regards,
Noob
↧
↧
Re: vsphere standard switch nic teaming
vSphere NIC teaming can work without the need of configure link aggregation on physical switches, but you will need to select the correct load balance algorithm, that should be different than the Route based on IP hash, which you already discovered that work only with link aggregation on physical switch side.
And the LACP works only with vDS, even on version 6.5, see: Host requirements for link aggregation for ESXi and ESX (1001938) | VMware KB
↧
vmnic buffer tx rx value cannot be changed
vmnic buffer tx rx value cannot be changed once the mtu is set to 9000. i.e RX preset maximum shows as 512 when the mtu is set to 9000. and if i reduce the mtu to 1500 then the rx preset maximum shows as 4096.
[root@localhost:~] ethtool -g vmnic2
Ring parameters for vmnic2:
Pre-set maximums:
RX: 512
RX Mini: 0
RX Jumbo: 0
TX: 4096
Current hardware settings:
RX: 512
RX Mini: 0
RX Jumbo: 0
TX: 1024
Any suggestions?
Thanks,
Suma
↧
Re: can't have a connectivity using a dvportgroup
Did you sending 802.1Q from VM's also can you check if DvSwitch has uplink connected and in teaming policy correct uplink is in active state?
-Saurabh Jain
↧