HP Procurve 2650 issue. 1

[khilton]

Hi all,
I am trying to get two HP ProCurve 2650's to work together, but I am not having any luck. I followed the manuals configuration by making one switch Master with static IP and auto join to yes. The second switch I made Candidate with auto join to yes as well. The first switch (Master) recognized the Candidate and I added it according to the manual. However, when I did add it to the stack all computers on the Candidate switch had no access to the network. can someone please tell me what I might be doing wrong?
Karsten

 

[raztaboule]

Perhaps a VLAN tag missing on the sport used for the stack ?

 

[khilton]

Would that get rid of my mismatch comment in the stack manager?

 

[khilton]

I managed to VLAN Tag port 47 on both switches. When I connected the switches and etc.. The second switch joined fine. I connected a couple of computers off of an older HUB and tested them on the switch with success. However, when I moved 10 more connections to the new switch I received an "Excessive broadcasts" error on the VLAN Tagged ports. So, I retagged port 50, one of the gb ports and tried there. then I received the error on that as well as on another port connected to a hub. I removed the hub and same thing still. any suggestions?

 

[krr]

Don't mess around with VLAN tagging if you don't need VLANs. Enable Spanning-Tree instead to prevent loops and the resulting broadcast storms which are the most likely cause for your network problems. I guess you have more than one connection between the two switches.

Stacking the 2600 switches is purely virtual and does not require any special or additional connections. Any regular Ethernet LAN connectivity will do. The stack commander communicates with stack members by means of Ethernet MAC address.

Hope this clarifies things a bit. *Rob

 

[khilton]

krr,
I appreciate the info. However, I have since resolved the issue. Apparantly, before I was hired, who ever set up the cables had two ports on the patch board connected to one cable, so when I hooked the two patch cables in from those two ports, it caused a big loop. The switches were not impressed by this. lol.
Thanks though.

 

[hinesjrh]

I believe I have a similar situation. I had three existing HP 2650 switches working together in a rack. One switch went bad, so I configured a spare switch I had (and I beleived the config. was exactly the same) but when we attempted to install it and plug in the patch cords it took down most of the network. So I had the tech. send me both switches so I could compare. The only difference I see in the configs. is as follows:

1) Relacement switch has a different snmp-server community name (old/bad switch was "public" but replacement has an actual name).

2) Replacement switch has command line "time timezone -5"

3) Replacement switch has command line "no timesync"

4) Old/Bad switch has command line stack join 00110a25a680, and replacement switch does not have a "stack join" command line.

What is the "stack join" anyway? What do I need to do to correct this problem (taking the network down when we plugged it into the other switches). I am a bit new to this, so any help is greatly appreciated.

FYI: The three switches in the rack are as follows:

Top switch is primary (has fiber connection), and switch below that is connected to switch above and switch below (via patch cables), and bottom switch (of three) is connected to both units above it via patch cable.