Hi, french & junior (vmWare) but still alive
Existing :
VMware PowerCLI 11.5.0 build 14912921
PowerShell 5/1/14409/1018
1Cluster/3ESXI/VCenter Vsphere 6.7u3d 15679281 (6.7.0.42200)
ESXi 1, 2 ,3 : Image Profile (Updated) DellEMC-ESXi-6.7U2-13006603-A00
Education environment.
What's going on :
Since november, one of the ESXi (but not always the same) lost sometimes connection with VCenter and/or other ESXi.
Over time, it went worst and worst. At this moment +/- 100 VMs were up.
December, a hundred more VMs have been created and put on line.
VMs when not on same EXSi can not communicate with each other, despite they use the same PG (and the exact same PG exists on the 3 ESXi, same name, same VLanID).
To get round this problem, I use VMs/Host rules, but it seems to overconsume network/vMotion and 'Connection lost' with ESXi appears more frequently...
Infernal spiral...
I don't know exactly which updates were done since it began to play up. Can I get a log somewhere ? Need ESXi to be update ? Which updates and how I can run them ?
Thanks.
Hi,
Can you insert Snippings of the network configuration of at least two Esxi hosts? (vSwitch, VDS, VSS, Port Group and VlanID).
Do some checks based on this article:
https://serverfault.com/questions/754429/guest-vms-cant-reach-each-other-on-different-esxi-servers
You have double checked the network parameters of the Esxi hosts, (Subnetmask, gateway, etc ...)
ARomeo
Hi,
Can you insert Snippings of the network configuration of at least two Esxi hosts? (vSwitch, VDS, VSS, Port Group and VlanID).
Do some checks based on this article:
https://serverfault.com/questions/754429/guest-vms-cant-reach-each-other-on-different-esxi-servers
You have double checked the network parameters of the Esxi hosts, (Subnetmask, gateway, etc ...)
ARomeo
Hi,,
I will use same test user & PG created with LucD here :Create vLans in personal folders with specific permission/role
The script created the same PG on the 3 ESXi (same name, same ID), on vSwitch2.
PS C:\script> Get-VirtualSwitch -Name vSwitch2 | Get-VirtualPortGroup | Format-Table -AutoSize
Name Key VLanId PortBinding NumPorts
---- --- ------ ----------- --------
GRETA-LAN-2137 key-vim.host.PortGroup-GRETA-LAN-2137 2137
GRETA-LAN-2137 key-vim.host.PortGroup-GRETA-LAN-2137 2137
GRETA-LAN-2137 key-vim.host.PortGroup-GRETA-LAN-2137 2137
Please, I need more help/accompany to get the configuration you ask me, I'm junior vmware...
Default TCP/IP Configuration Stack for each ESXi = 1 VMKernel Adapter, Gateway & DNS are the same & ok.
The 3ESXi + vCenter Server have good IP/netmask, on the same network range.
In fact, other VMs run with no problem here (DC, pfSense, VCSA, Veeam, Wsus,...) and are hosted in different ESXi and depending of vMotion moves.
ESXi are connected together through a physical switch.
Do no use route based on IP hash. Change that to originating virtual port ID.
I changed on the 3esxi but still get "Destination Host Unreachable" for my 2 test VMs (ping ok when VMs blocked on the same ESXi).
VLan were not known by the physical switches...