Weird Switching Issue

[Killer_Rubber_Ducky]

So, I am working on this weird Switching issue at a school and I need some advice.

The school recently added a Satellite Uplink/Downlink for their WAN. The link apparently made itself the ROOT bridge of the Main VLAN instead of the actual Management switch which is a Cisco Catalyst 3750. I advised changing the 3750 priority to 4096 to bring it back to ROOT bridge. Now, it lists as being ROOT on VLAN 1, 2, 3, and 12. There are 32 switches connected to this switch and all 32 switches are trunked on VLAN 1. The VTP has no Domain and no Domain password. At least one of the downstream switches is also trunked to another switch not connected to the 3750 also trunking on VLAN 1 which seems to me like it would cause a strange issue.
What should I do for this network to restore proper networking? My current thoughts are to redo the VLANs so that Everything trunked to the 3750 is using VLAN 99 for trunking and the VTP has a domain and domain password to prevent a rogue switch from joining the network as a server and taking everything over. Then, segmenting the ports out on the switches to using separate VLANs for different traffic. This would require reconfiguring ~ 32 switches manually and 1 of these switches is a HP switch while the rest are Cisco/Linksys switches with some being SMB switches that are limited in what you can do to them.

If I get the chance I can create a network diagram to better show the layout. But can someone tell me what I should be doing to fix the STP issue on the network for sure? I have had 4 semesters of Cisco CCNA prep school but we didnt really go in deep with STP.

Thanks.

 

[remixedcat]

What's the default route (VLAN) set to??

Also is it on a tagged or untagged port??

 

[Killer_Rubber_Ducky]

Everything seems to run over VLAN1 and VLAN1/VLAN 1 is the default VLAN and can not be renamed. Most of the SMB switches are using Tagged ports. The 3750 does not seem to have a web interface beyond a simple setup that provides next to no options.

 

[remixedcat]

What port and VLAN is the link on???

Is it acting as dhcp as well?

Please tell me more how its configed on the WAN side..

 

[Killer_Rubber_Ducky]

They are using Gigabit connections aka Gi1/0/1 to Gi3/0/*

I have not checked for DHCP yet. I go there today.

 

[remixedcat]

as far as I know any "satellite link" boxes are not supposed to even have router/switch functions and they are supposed to act like dumb modems and not be able to mess with VLANs/DHCP!

Unless there's an option to disable that. Why was it even set up that way?

If it REQUIRED to do so:

is yours doen like this example that I found really quick: http://www.freeccnaworkbook.com/workbooks/ccna/configuring-per-vlan-spanning-tree-protocol

there's root primary and root secondary and also "backup" You might want to make the link "backup" or "secondary"

Are you working with something like two management VLANs??

 

[Killer_Rubber_Ducky]

no just 1 managment VLAN. In fact, there is no "Management" VLAN. When I was remoted in, I had to create one. Everything was trunked over VLAN1 (default).

 

[Killer_Rubber_Ducky]

Here is the basic layout and issues at the network:

 

[remixedcat]

pirority should have worked. tell me more about this "satellite link" and what's up with it.

 

[remixedcat]

if not you may have to create another VLAN and tag and untagg appropriately.... without more info on that "satellite link" and how it works I have no idea what else....

 

[Killer_Rubber_Ducky]

well, we seem to have fixed it for the most part. We disabled one of the switches' Gi ports.changed Gi3/0/12 to VLAN 1, and disabled the port-priority mode. The root bridge is set correctly. Now, the issue seems to be with the WAPs. we can ping the gateway just fine over the LAN but when you go over WLAN, ping is fruity. See pings anywhere from 1.34ms to 18000ms as well as those pings to ping the WAP itself. Trying to figure that one out. Also having CPU load on the switches around 65% which seems kinda high but nothing like the 95% it was before.

Also found out that someone from Cisco supposedly came out and set this Charlie Foxtrot up. It seems sloppy since VTP and STP were set up when most of the downstream switches dont support VTP. Also how VTP domain is not set and no password on the VTP. STP is just screwy but works. 32 switches are trunked to 3 3750s stacked as 1. But 90% of the trunked switches dont support VTP. It seems odd that the Cisco guy said nothing about the conflict.

 

[remixedcat]

Cisco APs??

 

[Killer_Rubber_Ducky]

Cisco APs??

Mostly Ubiquity with a couple Engenius

 

[remixedcat]

That's really odd that just changing VLANs would affect the wireless that much... mabye the priority was helping??

Mabye something that was changed created a bottleneck. How are the APs connected and are they controller/controllerless?

 

[Killer_Rubber_Ducky]

eh, we gave up. We called in Cisco. It seemed fixed and then, it went back to the beginning again. So we called Cisco.

 

[remixedcat]

could it be the crappy satellite thingy??? what satellite thingy do you use??

 

[Killer_Rubber_Ducky]

it is not me. It is them. And, I am no longer there.

 

[remixedcat]

oh