Problem with certain ports on Catalyst 6506

Unanswered Question
Aug 20th, 2010

I am having a really strange problem with certain ethernet ports on my cisco catalyst 6506 switch.   I have four WS-X6148-GE-TX (48 port 10/100/1000 ethernet) cards residing in slots 3 thru 6.  I am having connectivity issues on ports 3/1, 3/26-30 and 4/26-30.  Any device plugged into these ports will have poor connectivity and a ping from any other device on the network (even plugged into the same switch) will result in about 500ms response time.  Strange thing is that it is ANY device plugged into these "bad" ports have the issue, so it's not a NIC driver or autonegotiation issue.   And the weird part is that its the pretty much exactly the same bad ports on the cards in both slot 3 and 4, so its most likely not a hardware issue, it would seem.   And the strangest symptom of all is that the issue only exists between 9am and 11pm.   Between 11pm and 9am, any devices on these ports respond to pings in < 1ms locally.   The supervisors are in slots 1 and 2 (WS-X6K-SUP2 with MSFC2),  What could possibly be causing these issues?

I have this problem too.
0 votes
  • 1
  • 2
  • 3
  • 4
  • 5
Overall Rating: 5 (1 ratings)
burleyman Fri, 08/20/2010 - 12:39

If I remember correctly that module  has ASCIS that are shared and they are in groups of 4 or 8 ( will have to track down the document) which so the Gigbit bandwidth is shared across those ports and if one port in that ASIC group is pushing alot of data the rest of the ports will suffer. Also the Buffer is size is only 1.4MB per 8 ports shared again. We had a very similar issue and we needed to upgrade to the 67xx series modules where each port has its own ASIC and Buffer. In the mean time you would have to take and isolate the heavy hitters to there own ASIC so it would leave port that you could not use in the ASIC group.



This Discussion

Related Content