05-08-2003 04:37 PM - edited 03-02-2019 07:13 AM
Hello,
we get on our Cat5500 the following messages:
%SYS-3-PORT_ERR:Port x/y swBusCRCErrorDrop
Concerning Cisco we should hard code duplex settings, but the are hard coded. Any ideas?
Regards, Snitram
05-08-2003 06:43 PM
What connects to these ports? Are both sides hardcoded? If it is a end workstation, you might want to check the NIC card/driver as well. Check out the following page on CCO
05-09-2003 02:17 AM
Hello,
there are workstations and a ISL-Trunk to a router. All ports are on all sides hardcoded. The error message is not constricted to one ore two ports, but they appear on all ports of the module.
Best regards, Snitram
05-09-2003 03:28 AM
Do you have any HUBs in your network? If so you need to make sure the duplex is set to half and not full. Also what type of cabling are you running across? If you are running on CAT5 or better speed should not be an issue, however, if you are running on CAT3, then the speed needs to be set to 10MB. Also, if you do have HUBs then make sure what speed they can run at, 10MB or 10/100MB, and that everything connected to them are set accordingly.
Hope this helps,
Burleyman
05-09-2003 03:48 AM
No Hubs. Cabeling: CAT5. Everything seems to be connected accordingly.
I have no other idea but hardware-issue, because the messages are comming from all ports of the module.
Best regards, Snitram
05-09-2003 04:10 AM
I did some research and did not find any known hardware issues. Check for any duplex mismatches on all ports especially ISL trunk ports. A mismatch on one single ISL trunk can cause this on other ports due to the following bug
http://www.cisco.com/cgi-bin/bugtool/onebug.pl?bugid=CSCdw11398
If you still have any issues, I suggest that you open a TAC case
06-03-2003 03:06 PM
Hi There,
Is this problem resolved? What actions you took?
We are having similar problem and almost did everything except changing the slot of the Ethernet card.
Thaks in advance.
Sunil
06-03-2003 11:27 PM
Hi Sunil,
after changing a defective 24-Port 10/100 MBit linecard the problem was resolved. But it was dificult to find out the linecard that was defective :-(
hth
Best regards, Snitram
06-04-2003 07:00 AM
Thanks Snitram,
I have already changed Ethernet blade (on the other end also). May be there is a problem in the backplane.
Thanks anyway.
Sunil
06-19-2003 07:43 AM
We replaced RSM/VIP combo, and have not received these errors in last 40 hrs. but now we are getting "dmaXferLengthErrors".
Sunil
06-19-2003 08:10 AM
Hi Sunil,
how often do you get this errors?
Do you have speed/duplex hard coded?
Best regards, Snitram
06-19-2003 09:20 AM
Hey Snitram,
We have received 8 "dmaXferLengthErrors" during last 40 hours after replacing the RSM/VIP. Speed/Duplex are already hardcoded on both ends.
Thanks
Sunil
06-19-2003 11:19 AM
Hi,
http://www.cisco.com/en/US/products/hw/switches/ps679/products_tech_note09186a0080143d13.shtml
Symptom 2: %SYS-3-PORT_ERR:Port 6/23 dmaXferLengthErrors (236)
Generally this message is logged when errordetection is enabled for port counters by issuing the command set errordectection portcounters enable. This command is only useful if a network impacting problem exists. By default, this command is disabled. The code dictates that these counters should be zero, but if they reach a certain threshold, you will off a syslog message. The oversubscription could also be due to a spanning tree loop which could have oversubscribed the port resulting in Direct Memory Access (DMA) queue being full. Disable the errordetection by issuing the set errordectection portcounters disable command and see if the message disappears. After disabling the errordetection counters if you still receive this message, open a case with the Cisco Technical Assistance Center (TAC).
Check the counters; if they do not increment, forget the only 8 messages.
Best regards, Snitram
Discover and save your favorite ideas. Come back to expert answers, step-by-step guides, recent topics, and more.
New here? Get started with these tips. How to use Community New member guide