6224 dont work stack with vmware 5.1

Hi,

Can you update the switch firmware to the latest version? There was a fix for some failover issues. http://www.dell.com/support/drivers/us/en/04/DriverDetails/Product/powerconnect-6224?driverId=77XG3&osCode=NAA&fileId=3288111910&languageCode=en&categoryId=NI

Are the ports in a static or LACP lag?

Hi Josh,

ok, im go update for lastest version, and about ports in a static or LACP lag, the ports are static.
but documentation of vmware and dell, with static ports the stack should work

regards

Hi Josh,

i update de firmware with lastest version, and the stack dont work in both switch. when i start/power on second switch 

the same show me in console the follow message and not work. ( UNIT 1 - Waiting to select management unit ) and not

happen.

if you can help me.

regards.

Are both units showing that they are unit one in the stack? Or does switch one work correctly and you only get an error on switch 2, if they are stacked all commands are run through the master. If you run show stack port what does it show?

Does show switch 2 from the switch 1 cli show the 2nd switch?

follow results of comands.

but the both units dont work togehter, 1 unit stay in standby mode.

dell#show switch

    Management Standby   Preconfig     Plugged-in    Switch        Code  

SW  Status     Status    Model ID      Model ID      Status        Version

--- ---------- --------- ------------- ------------- ------------- -----------

1   Mgmt Sw              PCT6224       PCT6224       OK            3.3.8.2    

2   Stack Mbr  Oper Stby PCT6224       PCT6224       OK            3.3.8.2    

and

#show stack-port 

                      Configured  Running

                        Stack      Stack      Link       Link

Unit    Interface       Mode       Mode       Status     Speed (Gb/s)

---- ---------------- ---------- ---------- ------------ ------------

1    xg1              Stack      Stack      Link Up      12         

1    xg2              Stack      Stack      Link Up      12         

1    xg3              Ethernet   Ethernet   Link Down    Unknown    

1    xg4              Ethernet   Ethernet   Link Down    Unknown    

2    xg1              Stack      Stack      Link Up      12         

2    xg2              Stack      Stack      Link Up      12         

2    xg3              Ethernet   Ethernet   Link Down    Unknown    

2    xg4              Ethernet   Ethernet   Link Down    Unknown

yes, each port of servers-esxi have lag configuration on each switch, but the traffic bettwen switch dont happen, the second switch dont work active-active with master, and i need config each port, jumbo frames in each switch, i cant config just one and replicate for other automatically, is this correct ??

regards 

That looks like the stack is operating correctly. Is each server in its own lag with one more on each of the switches?

Stacking makes the two switches act like a single switch, so if you need to configure ports on switch 2, it is done from the switch 1 cli, as it is the master,

http://www.dell.com/downloads/global/products/pwcnt/en/pwcnt_stacking_switches.pdf

ok, but the switch 1 i had her work in my enviroment, i add the 2 switch after. and for example, if i disconect the cable 1 port 1 of  switch 1, the enviroment
dont can stop, because the switch 2 and stack have work with this port and this not happen, i lost connection with this server. you can understand me ??

i need that 2 switch stay with HA, work active-active.

It sounds like it is not failing over to the 2nd nic port when the first one is disconnected, do you have a non vmware system with 2 nics that you can test the failover on? The issue might not be on the powerconnect side.

This is how it is set up on the Ethernet side?

Then if you break the connection to switch 1 it doesn't failover to switch 2. Are you disconnecting all of the cables from switch one, or shutting down those ports or just some of them to test?

If you don't have the switches stacked does it work properly?

exactly how it is the setup. if dont have the switches stacked. the enviroment work normal. i need this scenario stacked, because, if 1 port of server failure the traffic need be redirect for other switch with stack-HA and JOB cotinue working. In this scenario up the second switch stay and standby and not respond if 1 port of system failure.

regards.

The standby switch is the one in the stack that takes over in the event of the master failing, if you had a stack of 3 switches, you would have one master, one standby and one member. It is still participating in the stack. If you run initiate failover, it should switch which one is the master. Failover occurs naturally when a switch fails, not when a port does. So in your case disconnecting the cables from switch 1 is not going to cause it to failover to switch 2, as the switch is still communicating with the other switch through the stacking ports. Stacking is basically creating one big switch out of two or more switches all managed through a single switch CLI.

If 1/g1 and 2/g2 are in a LAG and 1/g1 goes down due to a disconnected cable, it will not failover the stack, it can still access 2/g1 through the stack. It seems like the distributed vmware switch is not failing over to send traffic to the 2nd switch if the primary nic link drops.

Is spanning tree enabled on the switches?

yes, spanning tree are enable on the switches.

neste caso eu não vou conseguir deixar os 2 switches ativos, e se uma porta falhar o outro switch continuar trafegando ??

Run a show spanning-tree active and make sure it isn’t discarding any links that it shouldn’t be, if the LAG is working properly it should treat it as a single link.