Issue with Connecting to ESXi 5.5 host using Linkagg

[laoistom]

Hi Folks,

I'll start by saying that I'm fairly new to Alcatel hardware but by no means am I new to networking on other platforms (cisco, juniper, hp, etc.....)

I recently had to setup two alcatel 6450 switches (One is a 24 POE the other a standard 24 port) in a stack and program it up for use in a data centre for Co-Location purposes.

At this stage we only have these two switches, a rack mounted Qnap and a Dell R730 running ESXi. I had absolutely no difficulty configuring the Qnap with a Link Aggregate connection (LACP LINKAGG 2 in the the config below). Once configured all aspects of connectivity worked exactly as expected including NFS mounting, CIFS shares, Failover, etc......

Now I also know that ESX and ESXi don't support Dynamic LACP, they only support static and I'm ok with that. So I configured a Static Linkagg (see config), configured vlan 2 port default 1 (to put my linkagg on native vlan 2) and then configured vlan 6 802.1q 1 to make the Linkagg accepts tagged packets on VLAN 6.

Now at this stage everything worked as you would expect for about 5 minutes. As we began to test remote connectivity we noticed that the management interface went offline on my laptop (connected to VLAN 2). I couldn't ping it at all, however my colleague sitting not 2 feet away from me also on VLAN 2 could ping it fine and also access the vSphere client without any difficulty. There were two VM's running on the host and while I couldn't ping the Management interface I could ping both of the hosts which were configured on VLAN 6.

As a test I connected myself to VLAN 6, I could still ping the two VM's but as before, couldn't ping the vSphere management interface. Meanwhile my colleague continued to have no issues what so ever. Eventually we tested if we could reach the management interface over the VPN we configured from our head office. We tested from two machines (windows 2003 and Windows 10) the windows 2003 could ping and connect to the management interface on esx but the windows 10 PC could not.

In the end we simply connected a third NIC on the host and configured that to vlan 2 port default 1/5 and set that as the management interface. As you might imagine this is not my preferred method of getting management working on an ESX host.

I've posted the config below in the hope that one of ye may be able to see something that my tired eyes have missed. Apologies for the long post but if you've read this far thank you for your patience and I hope that someone may be able to point me in the right direction.

************************************************************************************************************************************************

Welcome to the Alcatel-Lucent OmniSwitch 6450
Software Version 6.6.5.63.R02 GA, December 16, 2014.

Copyright(c), 1994-2014 Alcatel-Lucent. All Rights reserved.

OmniSwitch(TM) is a trademark of Alcatel-Lucent registered
in the United States Patent and Trademark Office.

-> write terminal
! Stack Manager :
! Chassis :
system daylight savings time disable
! Configuration:
! VLAN :
vlan 1 1x1 stp disable flat stp disable name "VLAN 1"
vlan 2 1x1 stp disable flat stp disable name "MANAGEMENT"
vlan 2 port default 1/5
vlan 2 port default 1/6
vlan 2 port default 1/7
vlan 2 port default 1/8
vlan 2 port default 1/9
vlan 2 port default 1/10
vlan 2 port default 1/11
vlan 2 port default 1/12
vlan 2 port default 1/13
vlan 2 port default 1/14
vlan 2 port default 1/15
vlan 2 port default 1/16
vlan 2 port default 1/17
vlan 2 port default 1/18
vlan 2 port default 1/19
vlan 2 port default 1/20
vlan 2 port default 1/21
vlan 2 port default 1/22
vlan 2 port default 1/23
vlan 2 port default 1/24
vlan 2 port default 2/1
vlan 2 port default 2/4
vlan 2 port default 2/5
vlan 2 port default 2/6
vlan 2 port default 2/7
vlan 2 port default 2/8
vlan 2 port default 2/9
vlan 2 port default 2/10
vlan 2 port default 2/11
vlan 2 port default 2/12
vlan 2 port default 2/13
vlan 2 port default 2/14
vlan 2 port default 2/15
vlan 2 port default 2/16
vlan 2 port default 2/17
vlan 2 port default 2/18
vlan 2 port default 2/19
vlan 2 port default 2/20
vlan 2 port default 2/21
vlan 2 port default 2/22
vlan 2 port default 2/23
vlan 2 port default 2/24
vlan 3 1x1 stp disable flat stp disable name "H004"
vlan 4 1x1 stp disable flat stp disable name "E001"
vlan 5 1x1 stp disable flat stp disable name "D008"
vlan 6 1x1 stp disable flat stp disable name "S008"
vlan 7 1x1 stp disable flat stp disable name "A012"
! VLAN SL:
! IP :
ip service all
ip interface "VLAN 1" address 10.3.1.254 mask 255.255.255.0 vlan 1 ifindex 1
ip interface "MANAGEMENT" address 10.3.2.254 mask 255.255.255.0 vlan 2 ifindex 2
ip interface "H004" address 10.1.2.254 mask 255.255.255.0 vlan 3 ifindex 3
ip interface "E001" address 10.3.4.254 mask 255.255.255.0 vlan 4 ifindex 4
ip interface "D008" address 10.3.5.254 mask 255.255.255.0 vlan 5 ifindex 5
ip interface "S008" address 10.3.6.254 mask 255.255.255.0 vlan 6 ifindex 6
ip interface "A012" address 10.3.7.254 mask 255.255.255.0 vlan 7 ifindex 7
! IPMS :
! AAA :
aaa authentication telnet "local"
aaa authentication http "local"
! PARTM :
! 802.1x :
! QOS :
! Policy manager :
! Session manager :
! SNMP :
! RIP :
! IPv6 :
! IP multicast :
! IPRM :
ip static-route 0.0.0.0/0 gateway 10.3.1.1 metric 1
ip static-route 10.3.2.0/24 gateway 10.3.2.254 metric 1
ip static-route 10.1.2.0/24 gateway 10.1.2.254 metric 1
ip static-route 10.3.4.0/24 gateway 10.3.4.254 metric 1
ip static-route 10.3.5.0/24 gateway 10.3.5.254 metric 1
ip static-route 10.3.6.0/24 gateway 10.3.6.254 metric 1
ip static-route 10.3.7.0/24 gateway 10.3.7.254 metric 1
! RIPng :
! Health monitor :
health threshold temperature 78
! Interface :
! Udld :
! Port Mapping :
! Link Aggregate :
static linkagg 1 size 2 admin state enable
static linkagg 1 name "S008-ESX"
lacp linkagg 2 size 2 admin state enable
lacp linkagg 2 actor admin key 1
static agg 1/2 agg num 1
lacp agg 1/3 actor admin key 1
static agg 2/2 agg num 1
lacp agg 2/3 actor admin key 1
! VLAN AGG:
vlan 2 port default 1
vlan 2 port default 2
! 802.1Q :
vlan 6 802.1q 1 "S008-ESX"
! Spanning tree :
bridge mode 1x1
! Bridging :
! Bridging :
! Port mirroring :
! UDP Relay :
! System service :
swlog console level info
! SSH :
! VRRP :
! Web :
! AMAP :
! Lan Power :
! NTP :
! RDP :
! VLAN STACKING:
! EFM-OAM :
! SAA :
! Loopback-detection :
! ERP :
! TEST-OAM :
! DHL :
! LLDP :
! DHCP Server :
! Stack Split-Protection Helper :
! Openflow :
->

[silvio]

Hi,
in my knowledge ESX doesn't support no static linkagg (like alcatel). This is loadbalancing between the two ports. So you have not to configure static linkagg. Try to use "normal" ports (with tagging). ESX will deside how to handle the packets.
regards
Silvio

[laoistom]

Thanks you for your really quick reply Silvio,

So are you suggesting I remove the Linkagg and run with a config like this :

vlan 2 port default 1/2
vlan 2 port defualt 2/2
vlan 2 802.1q 1/2
vlan 2 802.1q 2/2

Then on ESX would I leave the vSwitch in IP Hash mode or set it back to "Route based on Originating port ID.

Regards,

Laoistom

[silvio]

yes, you have understand correctly.
regards
Silvio

[laoistom]

Thanks for the advice. I've implemented that and so far it seems to be stable.

Cheers,
Laoistom

[devnull]

I remember of a customer running LACP with ESXI/VSphere on a VSphere distributed vswitch.

Afaik at that time setup only worked when using the (vsphere?) web-client, not the other tool.

But i can't remember that in detail. If you need it i can ask the customer whether he wants to provide a "Alcatel-LACP-ESXi howto"