Unsolved
This post is more than 5 years old
5 Posts
0
47993
May 5th, 2004 20:00
Unmanaged switch and Spanning Tree Protocol
I am trying to solve a rather strange problem. I have a client with a small network with a Powerconnect 2324. Internet connectivity is through DSL but there is also a VLAN to local hospital. The VLAN is controlled by a Hospital and connection is through a Cisco Switch of unknown model. There are 2 workstations with secondary IP addresses for the VLAN and static routes to the Cisco switch. The CIsco Switch is plugged in to the Dell Switch.
The problem is that on the weekends the Cisco Switch shuts down the link to the Dell switch with logs indicating that there is some kind of Spanning Tree Protocol loop. This is all I know from that end. It does not make sense to me but this is the data I have.
My question is: Is the Dell 2324 be sending any BPDU's? Does the Dell Switch do any STP at all????
Thank you,
DaleBJW
GregG1
2 Intern
•
812 Posts
0
May 6th, 2004 10:00
None of our unmanaged switches send BPDUs or participate in the Spanning Tree Protocol. Unless there is a redundant link between the Cisco and Dell switches, the only way the Cisco switch would be receiving an inferior BPDU would be through the connections on one or both of the workstations.
If the workstations are using two network interfaces, there may be some bridging being performed between the NICs.
DaleBJW
5 Posts
0
May 6th, 2004 11:00
I know. It does not make sense. There has to be information that I have not been given.
Thank you,
Dale
GregG1
2 Intern
•
812 Posts
0
May 7th, 2004 14:00
With one side set to auto-negotiate and the other hard-set to 100full, you will have a duplex mismatch and the link will be established at 100half. Unfortunately, this is simply the nature of Auto Negotiation.
DaleBJW
5 Posts
0
May 7th, 2004 14:00
I was back on site yesterday and the link was down again. We tested the cable between the Cisco switch and the Dell switch and it passed. Hospital MIS has a crossover cable connected to Cisco switch. Both switches were autonogating the connection. We set the Cisco switch to 100Mb full. My question is will the Dell switch have a problem with that? If it is trying to autonegotiate the speed and duplex, will it have a problem when the other end is set static?
Thanks,
Dale
sentinel-master
345 Posts
0
May 19th, 2004 10:00
Get access to the cisco switch and do "show log" or "show loggin buffered" and post results here
ALso a "sho cdp neigh" may give you topology information about the network - i.e who's attached where - i Suspect the Cisco switch is seeing its owns BPDU's coming back to it on another - so suspect you do have a physical L2 loop or patch somewhere causing this
DaleBJW
5 Posts
0
May 19th, 2004 13:00
Unfortunately, I have zero access to anything outside the immediate office. The office is stand alone with DSL for internet access. They then have this one data port connecting to main hospital via the Cisco switch. THere is no other physical connectivity other than through the Cisco switch.
We made two changes and the problem has gone away. 1) set the CIsco switch to 100 full, 2) locked down a wireless access point that existed prior to our coming in to solve the problem.
Something I suspected was that offices above or below could connect to the wireless access point and cause a loop. When you change two variables, who knows which one was correct. Any way something seems to have worked
Thank you all,
Dale