We have a 5 switch stack consisting of 2 each PowerConnect 5524 switches (one is master) and 3 each PowerConnect 5548 switches. Unfortunately we have encountered problems that we have attributed to the length of the HDMI cabling used in the stack (over 9M). We are planning to remove Stack Unit ID 5 and connect it using the Dell 10GB SFP+ to SFP+ 5M cable. Unit 5 has several ports configured as LAG members. The LAG configuration is part of the config file on the master. When I remove unit 5 from the stack I assume I will lose the LAG configuration and have to recreate it when it is a stand-alone with it's own config file. Can anyone confirm my assumption?
I doubt the config will be intact, unless it is the master or backup. To ensure proper configuration, I suggest starting fresh with the newly standalone unit. You can backup the config from the stack, and then copy the commands that are pertinent to this switch.
Daniel Covey Dell EMC | Enterprise Support ServicesGet support on Twitter: @DellCaresPRO
Download our QRL app: iOS, Android, Windows
Are the ports in each LAG spread out across the switches in the stack? Or all on one switch?
LAG members spread out:
LAG members on same switch.
All the LAG members are on the switch being removed from the stack.
Excellent, should be pretty straight forward to copy the pertinent lines from the current running configuration to the standalone unit.
So on this beautiful Sunday morning I am in the office reconfiguring the network switch stack. During the week I prepped by routing the 10GB SFP+ to SFP+ cable in place and a short HDMI cable to close the ring on switches 1-4. I configured LAG 1 for the domain controller and LAG 3 for one of the VM hosts on switch 4 and have temporary patch cables in place. The switch stack is configured for RADIUS so the AD-DC and RADIUS server (on the VM) need to be available in the stack or else I won't be able to login. I used the CLI and copied the running-config to startup-config and captured a backup via TFTP.First I connect the patch cables for the AD-DC and VM host. I disconnect the HDMI cables to remove Switch 5 from the stack and connect the short HDMI cable. I then connect the 10GB SFP+ to SFP+ cable to Switch 2 (TE2/0/1) and Switch 5 (TE5/0/1). I login to the stack and switch 5 and all the LAG configurations for that switch are gone. However, I notice I have no connectivity on the SFP ports. I notice that TE2/0/1 shows as 10G-FiberOptics. I notice in the port configuration documentation that I can't change the port from Fiber to Copper. So I check the manual and low and behold I find it states the ports are "10 Gigabit ports, designated as 1000Base-X-SFP+. The SFP+ ports are fiber transceivers designated as 10000 Base-SX or LX." Do you suppose that the Dell Technician that said the 10GB SFP+ to SFP+ Copper cable didn't know that the Dell PowerConnect 5548 doesn't support copper at all. I call into Tech Support but since we don't have ProSupport it's going to cost $599 to talk to someone or wait until between 8 and 5 tomorrow. Thanks Dell!In closing, all I can say is thank goodness I left the HDMI cables in place and didn't modify the config files. I simply removed the short HDMI cable and reconnected the longer runs to add Switch 5 back into the stack. I'll deal with this tomorrow.
Sorry to hear you ran into some issues.The SFP+ ports support the following transceivers:
SFP SXSFP LXSFP 1000Base-TSFP+ SRSFP+ LRSFP+ LRMSFP+ direct attach cable(PASSIVE)What is the part number for the transceiver/cable you were attempting to use?
470-AAVG, Dell Networking, Cable, SFP+ to SFP+, 10GbE, Copper Twinax Direct Attach Cable, 5 Meter,Customer Kit
That cable is supported on the 5500 series switches. The cable includes built-in transceivers, there is no need to change the port from fiber to copper. You can test connectivity by performing a loop test, plugging the cable into both SFP+ ports on the same switch. The lights on each port should come on.
Some transceivers are only supported on newer firmware. If the firmware is older, I suggest updating and testing the connection again.
The link lights come on in the loopback so I connected the SFP+ cable back into the same ports on Switch 2 and 5 but left the stacking cables in place and have an established link and the operational status in the GUI shows active. So the rabbit hole is getting deeper. Why when the switch was removed from the stack didn't I have a link?
Glad to hear the cable does work. It is kind of hard to speculate why it would not work when the switch was standalone.
When you are able to test this again, I suggest running some of the following commands to gather some further details.
console# show interfaces statusconsole# show spanning-treeconsole# show interfaces countersconsole# show logging
Also, double check the port configuration to ensure it is the same as when it is in the stack.
If the port is showing as administratively down, then issue the command no shut, on the interface.
So here's more to ponder. Why would Auto Negotiation be greyed out on TE2/0/1 but editable on TE5/0/1 and why would I not be able to edit the MDI/MDIX settings in the GUI on either (current is MDI)?
I am not certain why it is grayed out. What happens when you issue the following command on 2/0/1
Console(config)# interface tengigabitethernet 2/0/1
MDIX is not applicable to fiber connections.