Tek-Tips is the largest IT community on the Internet today!

Members share and learn making Tek-Tips Forums the best source of peer-reviewed technical information on the Internet!

  • Congratulations biv343 on being selected by the Tek-Tips community for having the most helpful posts in the forums last week. Way to Go!

Very Crazy Problem

Status
Not open for further replies.

MrNick0483

IS-IT--Management
May 12, 2008
151
US
Here goes....

I currently have a network filled with 3560's across 3 locations with a 4510R in the core doing all or my layer 3.

STP is enabled
All switches are in a passworded VTP domain.

The problem is there are five offices located 3 side by side, and two straight across the hall way. At the same time every sigle day for the last 4 days until later on that evening (3:30pm - 9:00 ish) it is like someone shutdown these five switch ports, and the 5 workstations do not show a link on the switch. Each drop from the IDF to each office has been visually traced, and recertified up to gig speed, eliminating the cable being the problem. Each drop is located inside a cable tray away from any RF.

If you move the drops to a different port the link still does not come up.

I am open for many suggestions.

 
Are these offices in the same VLAN? Trying to determine if you are losing one or multiple VLANs. Also ensure that you have all ports connecting to hosts configured as access ports with portfast and BPUGuard. Also if these are in the same VLAN is there anything else that is in that VLAN? Is it disconnecting as well?

B Haines
CCNA R&S, ETA FOI
 
Yes this entire buidling is in the same VLAN along with each port being configured as an access port with portfast enabled. There are other offices in the same building on the same VLAN and same switch that keep right on working, while the others are out.
 
I recently read a post where a user had a very similar problem---solved it by uninstalling XP SP3. The computers had recently updated to XP service pack 3...

Burt
 
Are the pc's going to sleep? Any recent updates at all? Any chance of swapping one with one that does not shut off, to narrow it down to the pc(or eliminate the pc)?

Burt
 
The PC's are wide awake, i even reimaged one to eliminate that software issue.
 
I would patch a working one that is alive on the network at the time into one of the ports on the switch that is supposed dead and verify if its the switch or the pcs' . Switches operate on asics which control groups of ports usually 8 or 16 ports and its possble one asic is flakey though having happen at exactly the same time every day I would think would rule that out .
 
WOuld it matter that around this time this is the momst high traffic point of the day?
 
Are you trunking all VLANs everywhere? Is it possible you have either loads of broadcasts or unicast flooding and you are just overwhelming your access ports?

It might be worth explaining the topology in a bit more detail. I have seen entire networks brought to their knees due to unicast flooding (which was due to a fundamentally bad design).

Andy
 
Shouldn't matter what time of day , you can look to see what every port is doing on the 3560's utilization wise at a given time just use the "show controllers utilization " command and it gives port utilization and even switch fabric utilization .
 
Yes, and if you find that there is a crazy amount of broadcast traffic floating across your network you may consider another VLAN or two...

B Haines
CCNA R&S, ETA FOI
 
First thing is to narrow it down to the pc or the switch/traffic.

Burt
 
Status
Not open for further replies.

Part and Inventory Search

Sponsor

Back
Top