I have a 2 C3750G-12S's in a stack running c3750-ipservicesk9-mz.122-44.SE. On one of the 3750's some (not all) of my interfaces are showing an excessive number of symbol frame errors and the interfaces are bouncing up and down from time to time. On the other switch in the stack this does not happen. I have tried swapping out all hardware, patch cables, etc with the same result and had the fiber tested as well. The only thing I can find is that the problem is specific to this switch. I opened a case with Cisco a couple months ago about this and they sent a replacement switch, however the problem still seems to be occuring.. Any ideas on what else could be causing this?
there waqs another thread about symbol errors
october 9 2008 Lan switching and routing
I'm not able to attach the link.
here is the answer
Posted by: rmcarthur - Hot Staging Engineer, Getronics UK Ltd - Oct 9, 2008, 5:02am PST
Gig ethernet uses 10 bit encoding instead of 8 bit. These 10 bit codes are referred to as symbols. Symbol errors can occur due to dirty fibre, noisy lines or, rarer, a hardware problem.
Unless they are increasing fairly quickly or regularly they are not a cause for concern.
Clear the counters, check your connections and monitor the situation.
Hope this helps.
Hope to help
And they is the indication of everything I have read. However if I take the exact same fiber, patch cables, gbic and put it into a different switch. The errors go away?
I've read again your first post more carefully, you are probably the one that knows symbol errors better.
You have changed everything : the fiber cables, the GBICs, even the switch and the problem is still there.
I know that it can be a problem, what if you disconnect the stack and use it standalone ?
In all your tries the other switch in the stack is always there.
Is the complaining switch a slave in the stack or it is the master ?
I'm proposing this because we had a strange problem with a stack: after few hours of uptime it was impossible to telnet to the master and even for the master was impossible to setup any type of TCP session. We had a case for two mounths.
At the end after a failed IOS upgrade we decided to remove the stack and we solved the issue.
Hope to help
I have exact the same problem with a C3750G-12S: Symbol errors on 6 ports, the ther 6 orts are fine.
Did you found the solution of your problem ?
I will replace the switch...
I already replace one 3750-12 a couple of month ago with a similar problem.
Could thie be an hardware related issue with the C3750-12 !
Transmit GigabitEthernet1/0/1 Receive
3307364551 Bytes 1877476777 Bytes
53434174 Unicast frames 44157108 Unicast frames
3656884 Multicast frames 640584 Multicast frames
577233 Broadcast frames 152088 Broadcast frames
0 Too old frames 1754101839 Unicast bytes
0 Deferred frames 95425267 Multicast bytes
0 MTU exceeded frames 21331158 Broadcast bytes
0 1 collision frames 0 Alignment errors
0 2 collision frames 61 FCS errors
0 3 collision frames 0 Oversize frames
0 4 collision frames 0 Undersize frames
0 5 collision frames 47 Collision fragments
0 6 collision frames
0 7 collision frames 14704371 Minimum size frames
0 8 collision frames 7114065 65 to 127 byte frames
0 9 collision frames 14719086 128 to 255 byte frames
0 10 collision frames 4648391 256 to 511 byte frames
0 11 collision frames 999915 512 to 1023 byte frames
0 12 collision frames 2775958 1024 to 1518 byte frames
0 13 collision frames 0 Overrun frames
0 14 collision frames 0 Pause frames
0 15 collision frames
0 Excessive collisions 14882 Symbol error frames
0 Late collisions 112 Invalid frames, too large
0 VLAN discard frames 0 Valid frames, too large
0 Excess defer frames 2872 Invalid frames, too small
5819537 64 byte frames 0 Valid frames, too small
11782336 127 byte frames
13429217 255 byte frames 0 Too old frames
3049344 511 byte frames 0 Valid oversize frames
12077112 1023 byte frames 7 System FCS error frames
11510752 1518 byte frames 0 RxPortFifoFull drop frame
0 Too large frames
0 Good (1 coll) frames
0 Good (>1 coll) frames
In my experience symbol errors are usually caused by a physical layer problem. If you have 6 ports that have errors and other ports that are running clean then an easy test would be to swap connections (including patch cable, GBIC, etc) between a good port and a port with errors. If the errors stay with the port you do have a switch problem. If the errors move with the cable then it is a problem in the cabling/connector infrastructure.
thanks for the reply.
I replaced the 3750-12 switch this morning using the same cables and SFPs and now I have no errors.
I am now convinced that there is an HW issue with some 3750-12's ! I have seen in this forum that I am not the ony one that encountered this problem.
Thank you for posting back with your results. If you use the same fiber and same patch cables, replace the switch and the errors stop then it is a pretty clear demonstration that this time the error was caused in the switch.
It makes the forum more useful when people can read about a problem and can read what was done to successfully resolve the problem. Thank you for providing that follow up.
I wonder if this isn't some sort of design flaw or bug or... I dunno.. but I too have a replacement switch here.. I also replaced the switch 3-4 months ago due to the same issue... Thats why I was puzzled.. Shouldn't have to keep replacing switches like this.. Unless it is part of an initiative to keep people on smartnet
experience/information sharing, this is the power of such a forum ! I hope (as already happened) that other people that encountered the same problem will inform all using this forum.
I hope these messages will also help Cisco to investigate in detail this issue.