Need help with 6500 Errors

[ralphcarter]

We just installed a brand new 6513-E switch with sup-720-3B

I keep getting these errors in the log. I have an open TAC case but no response to a fix yet.

If anyone got this error before, please let me know.

Dec 23 19:50:54.614: %SIBYTE-CFC2-4-SB_LATE_COLL: Late collision on mac 1 (count: 1174)

Dec 23 19:55:40.250: %SIBYTE-CFC2-4-SB_LATE_COLL: Late collision on mac 1 (count: 1175)

Dec 23 19:56:19.210: %SIBYTE-CFC9-4-SB_LATE_COLL: Late collision on mac 1 (count: 1205)

Cisco IOS Software, s72033_rp Software (s72033_rp-IPSERVICESK9_WAN-M), Version 12.2(33)SXI5, RELEASE SOFTWARE (fc2)

cisco WS-C6513-E (R7000) processor (revision 1.0) with 458720K/65536K bytes of memory.

Processor board ID xxxxxxxx

SR71000 CPU at 600Mhz, Implementation 0x504, Rev 1.2, 512KB L2 Cache

Last reset from s/w reset

6 Virtual Ethernet interfaces

48 FastEthernet interfaces

254 Gigabit Ethernet interfaces

1917K bytes of non-volatile configuration memory.

8192K bytes of packet buffer memory.

65536K bytes of Flash internal SIMM (Sector size 512K).

Configuration register is 0x2102

[johnlloyd_13]

kindly hardcode speed/duplex settings your ethernet ports.

[sdheer · Accepted solution]

%SIBYTE-CFC2-4-SB_LATE_COLL: Late collision on mac 1 (count: XXX) indicates
a late collision on the EOBC channel has been detected on line cards.
The EOBC channel is a shared media used by the line cards to exchange
information with the Supervisor engine (like CEF tables, statistics,downloading code, etc). This is why you may see several line cards reporting
this problem at about the same time.

From the line cards you can confirm this is happening using the "remote
command module  show eobc" and check if the counters rx_bad_pkts,
rx_fcs_err_pkts and
tx_bad_pkts increment.

From the outputs you provided we see that for whatever reason line card in
slot 2 is sending a lot of bad packets to the EOBC channel which is
congesting the media, so I would recommend the following:

1) Reseat module in slot 2 into the chassis and see if the
SIBYTE-CFCx-4-SB_LATE_COLL stops being reported by the system. Also check
with the show eobc command described above if the errors stop incrementing
by issuing the command 30 seconds away from each other.

2) If the problem persist, reseat the module in slot 1 into the chassis and
see if the problem is cleared using the same logic as in 1).

3) If problem persist at this point, repeat the process for other slots as well and see if that solves your issue.


Swati

** Please let us know and rate if you find it helpful **

[ralphcarter]

Thanks I will try this.